Review of Devonian-Carboniferous Boundary sections in the Rhenish Slate Mountains (Germany)

Thirty Devonian-Carboniferous Boundary sections of the Rhenish Slate Mountains and adjacent subsurface areas are reviewed with respect to litho-, event, conodont, ammonoid, sequence, and chemostratigraphy. In the interval from the base of the uppermost Famennian (Wocklum Beds, Wocklumian) to the base of the middle Tournaisian (base Lower Alum Shale), 11 conodont and 16 ammonoid (sub)zones are distinguished. The terminology of the Hangenberg Crisis Interval is refined, with an overall regressive Crisis Prelude below the main Hangenberg Extinction, which defines the base of the transgressive Lower Crisis Interval (Hangenberg Black Shale). The glacigenic and regressive Middle Crisis Interval (Hangenberg Shale/Sandstone) is followed by the overall transgressive Upper Crisis Interval that can be subdivided into three parts (I to III) with the help of conodont stratigraphy (upper costatus-kockeli Interregnum = upper ckI, Protognathodus kockeli Zone, and lower part of Siphonodella (Eosiphonodella) sulcata s.l./Pr. kuehni Zone). Protognathodus kockeli includes currently a wide range of forms, which variabilities and precise ranges need to be established before a precise GSSP level should be selected. Returning to its original definition, the former Upper duplicata Zone is re-named as Siphonodella (S.) mehli Zone. It replaces the S. (S.) jii Zone, which is hampered by taxonomic complications. The S. (S.) quadruplicata Zone of Ji (1985) is hardly supported by Rhenish data. The entry of typical S. (S.) lobata (M1) characterises an upper subdivision (subzone) of the S. (S.) sandbergi Zone; the new S. (S.) lobata M2 enters much earlier within the S. (S.) mehli Zone. The ammonoid-defined base of the Wocklum-Stufe (Upper Devonian = UD VI) begins with the Linguaclymenia similis Zone (UD VI-A1). The oldest S. (Eosiphonodella) enter within the Muessenbiaergia bisulcata Zone (UD VI-A2). The traditional Parawocklumeria paradoxa Zone of Schindewolf (1937) is divided into successive P. paprothae (VI-C1), P. paradoxa (VI-C2), and Mayneoceras nucleus (VI-C3) Subzones. In the lower Tournaisian (Lower Carboniferous = LC I), the Gattendorfia subinvoluta Zone is subdivided into G. subinvoluta (LC I-A2) and “Eocanites” nodosus (LC I-A3) Subzones. The Paprothites dorsoplanus Zone (LC I-B) can be divided into Pap. dorsoplanus (LC I-B1) and Paragattendorfia sphaeroides (LC I-B2) Subzones. Potential subdivisions of the Pseudarietites westfalicus (LC I-C) and Parag. patens Zones (LC I-D) are less distinctive. The unfossiliferous or argillaceous upper part of the Hangenberg Limestone and the overlying Lower Alum Shale Event Interval remain regionally unzoned for ammonoids.


Introduction
The Rhenish Slate Mountains, especially the northern and eastern Sauerland in the east of the river Rhine, are the most important of the classical regions for the study of the Devonian-Carboniferous Boundary (DCB) and of the global mass extinction interval that is now known as Hangenberg Crisis (e.g. Kaiser et al. 2015). Following a proposal by Paeckelmann and Schindewolf (published in 1937), the DCB type section was defined during the Second Heerlen Congress in 1935 in a railway cut near Oberrödinghausen. This was the first international chronostratigraphic boundary that was established at all. Unfortunately, it had to be discarded when a hiatus was discovered right at the chosen boundary level . During the following search for a new stratotype, many old and new lateral sections were investigated in detail. Results were published in a series of monographs, review papers, and field guidebooks. Important summaries are, for example, Stoppel (1977), Paproth andStreel (1982, 1984), Paproth (1986), Paproth et al. (1986b), , Streel et al. 1993, Korn et al. (1994a), , , and Becker et al. (2016b). There is a flood of individual papers (e.g. many contributions by D. Korn, F.W. Luppold, and co-authors), which are fully referenced in the individual section descriptions. These form the fundament of our review, which includes a smaller amount of unpublished information. Since research is ongoing, this paper will be completed soon by forthcoming publications. For example, a new study will provide details of the currently best Rhenish DCB section at the Borkewehr near Wocklum (northern Sauerland). A palaeogeographical overview was given by Paproth et al. (1986a), with recent updates by Koltonik et al. (2018).
For details concerning the global, first order Hangenberg Crisis or mass extinction, the reader is referred to the review by Kaiser et al. (2015). Becker et al. (2016a) summarised the DCB bio-and event stratigraphy, with the Rhenish succession as a sort of "standard reference" for global correlations. This approach is elaborated here by the introduction of a new formal subdivision and nomenclature for the distinctive intervals of the Hangenberg Crisis. We supply a summary of Rhenish DCB conodont and ammonoid zones, with new subzones, and the evidence they are based on. Due to the considerable number of Rhenish DCB sections and the wealth of literature, our review will exploit the extensive published evidence by quoting for each section all relevant papers, and by focusing on the staged entries (FADs, local FODs) and disappearances (LADs/local LODs) of stratigraphically relevant taxa and rare/unusual taxon occurrences. For detailed data on sedimentology and the complete faunal records, the reader should survey the quoted papers. In the section discussions, we emphasise aspects that are of especial importance for the current DCB revision and selection of a future GSSP level and section. Since significant parts of the reviewed information have previously been published in German, our paper aims to recycle the incredibly rich Rhenish DCB record for a wide international audience. limestones of the Aachen and Velbert regions, which are, unfortunately, mostly very poor in conodonts. The zone definitions and terminology used here follow, with one exception at the main Hangenberg Extinction, the FADs of name-giving index species. The upper boundaries of zones/subzones are automatically defined by the base of the next higher zone/subzone. Besides the defining index species, there are other marker species, which have their FADs within given zones, but not necessarily precisely at the same level as the index species. For the literature of given reference beds, see the individual section descriptions. Examples of marker conodonts are given in Fig. 2.
In the Rio Boreado section of the Carnic Alps, Bi. ultimus ultimus, Pa. gracilis gonioclymeniae, and Ps. marburgensis trigonicus enter synchronously ). In other Carnic Alps sections and in the Rhenish Massif, the situation is much more complex. Pseudopolygnathus marburgensis trigonicus seems to enter first (base of Wocklum Limestone at Oberrödinghausen, Schindewolf-Bed 23, Bed 181b of Sacher 2016;Sample 11 of Eickhoff 1973), followed by Pa. gracilis gonioclymeniae and Br. suprema at the level of "Sphenoclymenia" cf. brevispina (Oberrödinghausen, Schindewolf-Bed 22, Kürschner et al. 1993;Bed 185b of Sacher 2016), and then by the name-giving zonal species (Müssenberg, Bed 80, Luppold et al. 1994;Oberrödinghausen, Schindewolf-Bed 20 = Bed 191b of Sacher 2016). It is well possible that more detailed sampling will change this picture. Currently, it is not clear whether the Rio Boreado or Rhenish patterns represent better the FADs of the three marker species.
The oldest Rhenish Pr. meischneri, which is the oldest member of Protognathodus, occurs at Oberrödinghausen in Bed 20 , slightly below the first Kalloclymenia. The taxon is much too rare in the Wocklum Limestone to have biostratigraphical significance. As outlined by Tragelehn (2010), Becker et al. (2013), and Söte et al. (2017), there are early relatives of Siphonodella, which represent two different genera. These "siphonodelloids" have Polygnathus-type upper platforms, for example as in Po. inornatus and Po. symmetricus, but nonpolygnathid, extended basal cavities. The group requires detailed taxonomic treatment (see Tragelehn 2010;Becker et al. 2013). Both the Po. inornatus-symmetricus Group (see Corradini et al. 2016) and the "siphonodelloids" enter above the base of the Bi. ultimus ultimus Zone and specific forms may enable a future subdivision.
Subdivision: In the Müssenberg type section, Bi. muessenbergensis appears high in the S. (Eo.) praesulcata Zone (Bed 18, Luppold et al. 1994), at the level of the May. nucleus Subzone (UD VI-C 3 ). This may provide a correlation level within the poorly calcareous "Wocklum Beds" of Stockum (Bed 172, Clausen et al. 1994). The oldest known Pr. collinsoni of the Rhenish Massif was listed together with various pre-Crisis palmatolepids, bispathodids, and Ps. marburgensis trigonicus from Seiler, Section I (Sample 1, . This corresponds to an occurrence at the top of the S. (Eo.) praesulcata Zone at Grüne Schneid in the Carnic Alps (e.g. Kaiser et al. 2006;Corradini et al. 2017;Schönlaub 2018). Unfortunately, the species is too rare to be used as a zonal index form.
Discussion: Since the LOD of Pa. gracilis gonioclymeniae varies widely between sections, and even between different sampling campaigns of the same section, Kaiser et al. (2009) discarded the Middle praesulcata Zone and replaced it by the ckI, which was defined by the synchronous LOD of many taxa, such as Bi. costatus, Bi. ultimus ultimus, Ps. marburgensis trigonicus, Pa. gracilis gracilis, Pa. gracilis expansa, and Pa. gracilis sigmoidalis. Whilst these do not re-occur in the Upper and post-Crisis Intervals of calm pelagic carbonate platform settings ), reworking and downslope transport at seamounts, or stratigraphic leak due to the bioturbation of extremely condensed successions, may lead to post-crisis occurrences, always in very low numbers, including various Rhenish sections. Their reworked nature has been questioned by several authors but the assumption of a survival without recovery and without a new spread into the easily accessible, stable platform settings is more unlikely.
In the Rhenish Massif, the ckI is mostly represented by siliciclastics that have not yielded any conodonts so far. Therefore, the regional ckI conodont record is very poor, with the exception of some limestone nodules from near its top at Drewer (Bed 97, Korn et al. 1994b). They yielded an assemblage with Pr. meischneri, Pr. collinsoni, and Neo. communis communis, which are common in ckI faunas of many other regions (see compilation in Becker et al. 2016a, p. 363). Since there are no newcomers in such faunas on a global scale, it is currently not possible to replace the ckI by a FAD-based zone.
Several attempts to retrieve conodonts from weathered Hangenberg Black Shale failed. So far, no conodonts have been observed on black shale bedding surfaces. It is theoretically possible that some victim taxa survived into the conodont-poor Lower Crisis Interval, similar as postclymeniids did among the ammonoids (see Corradini and Spalletta 2018).
Protognathodus kockeli Zone  Synonyms: Upper praesulcata Zone of Ziegler and Sandberg (1984a), lower part of extended Pr. kockeli Zone in Corradini et al. (2016). Reference sections: Neu-Moresnet (Reissner 1990), Oese (Bed A,, Apricke , Effenberg (3 cm calcareous shale with Guerichia to ca. Bed 18c), Hangenberg (Sample I of Luppold et al. 1994), Drewer (Beds 100 and 1), Wocklum , Stockum , Eulenspiegel (Bed 21), Scharfenberg   . A post-Crisis spread of pseudopolygnathids is more characteristic for the basal Hangenberg Limestone of other sections (see Scharfenberg). Discussion: The precise use of the zone for a future DCB GSSP definition is hampered by a wide and variable definition of the species, which is based on intraspecific variability and rapid evolution of Protognathodus in the early recovery phase of the global Hangenberg Crisis. The types of Pr. kockeli come from the Wocklum or Borkewehr section, which is currently under re-study Hartenfels 2017 Hartenfels et al. 2017a, b;Hartenfels and Becker 2019). The holotype and figured paratypes of Bischoff (1957) represent advanced forms with either two rows of nodes on each platform side or incipient second rows (consisting of two nodes only) on one or both sides. In the first limestones above the Hangenberg Sandstone, there are much simpler forms, partly with only one row of nodes on one side and with at least three nodes defining a row. "Double-rowed" forms sensu Kaiser et al. (2019) possess a row of nodes on each platform side. Intermediates from Pr. collinsoni display at least on one side two nodes that are arranged in parallel to the carina. Complexly ornamented collinsoni variants possess many randomly distributed nodes. There are also Pr. kockeli morphotypes that are intermediate towards Pr. kuehni, (Figs 2e-h), which is characterised by transverse ridges. The total Protognathodus variability of the Stockum Limestone is even higher and includes aberrant morphotypes (e.g. specimens identified as Pr. praedelicatus, Pr. cordiformis, Protognathodus sp. A and B in Luppold et al. 1994). Kaiser et al. (2019) showed that platform shape (e.g. asymmetry) and blade curvature are less useful to define specific Protognathodus morphotypes but these features require further evaluations (Kaiser and Hubmann, this vol.). The precise succession of morphotypes and their ontogeny and variability have to be worked out before protognathodids can be used as a DCB GSSP criterion.
Polygnathus purus purus enters at Wocklum in the second bed of the Pr. kockeli Zone (Bed 4b, Hartenfels et al. 2017a). However, Kaiser (2009) noted a much earlier occurrence in the Montagne Noire, from Bed 70 (lower ckI) of the current DCB GSSP at La Serre, which is the bed just above the supposed HBS equivalent (Bed 69;see Feist et al. 2020). Supposed pre-Crisis occurrences reported by Corradini et al. (2003) have not been illustrated. Close relatives of Po. purus purus characterised by a more anterior position of the basal pit were described from the upper ckI of southern Morocco (Becker et al. 2013) and identified as Po. cf. purus purus. They suggest that it is possible to distinguish very early and typical forms of the subspecies.
Reference sections: Hasselbachtal (Beds 84N-81N), Oese , Apricke , Oberrödinghausen (Beds 6A-5c, Voges-Samples 1-3), Effenberg (ca. Beds 18d-l), Hangenberg (Sample II of Luppold et al. 1994), Müssenberg (Beds 3a-3B), Stockum (Beds 100-97, especially Unit Alb-C-c, Sample 330 of Alberti et al. 1974), Drewer , Eulenspiegel (Bed 22), Scharfenberg . Other index species: At Stockum, Bed 100 yielded Ps. fusiformis . The alternative shallow-water siphonodellids of South China (e.g. S. (Eo.) homosimplex, S. (Eo.) semichatovae) have not yet been found in Rhenish sections. The same applies to Clydagnathus-Patrognathus faunas that appear in the basal Tournaisian of Russian and Chinese sections (e.g. Becker et al. 2016a;). Discussion: Kaiser and Corradini (2011) documented the currently subjective identification of S. (Eo.) sulcata in many papers. Many specimens do not conform to the type North American lost specimen, which came from the basal Tournaisian Henryville Bed within the Ellsworth Member of the New Albany Shale (Evans et al. 2013). Until further revision, curved and ribbed basal Carboniferous siphonodellids are referred to as S. (Eo.) sulcata s.l. Following Kaiser and Corradini (2011), it is their Morphotype 5 that has the best potential to define a refined S. (Eo.) sulcata Zone. A joint S. (Eo.) sulcata s.l./Pr. kuehni occurrence in the top Stockum Limestone at Seiler, Section I (Bed 5), resulted in a first close alignment of both FADs ). Based on direct co-occurrences in the Graz Palaeozoic and Carnic Alps, Kaiser et al. (2019) confirmed the usefulness of (typical) Pr. kuehni as an alternative index species for the S. (Eo.) sulcata (s.l.) Zone. This is especially important for the majority of sections, where siphonodellids are absent in the Upper Crisis Interval (Stockum Limestone and equivalents, see compilation of Kaiser et al. 2015, tab. 5). It results in a combined S. (Eo.) sulcata s.l./Pr. kuehni Zone that ranges from Upper Crisis Interval III into the basal post-crisis time (basal Hangenberg Limestone). However, revisions of Protognathodus taxonomy, especially of the complex Pr. kockeli-kuehni transition (see Kaiser et al. 2019), have to consider the fact that the types of Pr. kuehni are not from the S. (Eo.) sulcata s.l./Pr. kuehni Zone, but from the younger S. (S.) duplicata Zone at Seiler, Trench 2.
In order to avoid these taxonomic complications, the jii Zone of Becker et al. (2016a) is replaced by the re-named S. (S.) mehli Zone, returning to the original definition of the Upper duplicata Zone by Sandberg et al. (1978). However, it is unfortunate that S. (S.) mehli (= cooperi M1) is relatively rare in several sections of the Rhenish Massif.
Subdivision: Sandberg et al. (1978) suggested that S. (S.) carinthiaca and S. (S.) cooperi s. str. (= Morph 2) enter later than S. (S.) mehli (= S. cooperi M1). Published faunas from Rhenish sections do not provide any support. As explained below, the FAD of S. (S.) quadruplicata occurs in several sections in the S. (S.) mehli Zone or even before (e.g. Seiler Trench 2, Bed 14, Koch et al. 1970), well before the entry of S. (S.) sandbergi or S. (S.) lobata. Currently, it is unclear whether all or only some of these records refer to the early M2 in the sense of the S. (S.) jii lectotype. Sandberg et al. (1978) showed that the FAD of S. (S.) obsoleta lies near the top of the Upper duplicata Zone (e.g. Oese, Kaiser-Bed 22b = R; Seiler sections, compare range in Ji 1985). The species occurs at Oberrödinghausen well below the base of the sandbergi Zone (Bed 3d 2 = Voges 1959, Sample 6) and jointly with the first S. (S.) belkai M2 . The FADs of both taxa mark in the Rhenish Massif a position in higher parts of the S. (S.) mehli Zone but currently we refrain from proposing a new subzone.
Since Voges (1959), various authors correlated the triangulus triangulus Zone with a high level in the Hangenberg Limestone (e.g. Ziegler 1971). However, there is evidence from several sections (Drewer, Seiler sections) that the index species may enter much lower, around the base of the S. (S.) mehli Zone. For example, Koch et al. (1970) recorded Ps. triangulus triangulus at Seiler Trench 2 from Bed 13, below the FAD of S. (S.) obsoleta. Ziegler (1971) marked at Oese the base of the triangulus triangulus Zone at the base of Bed M, within the range of Paprothites (Korn and Weyer 2003). A possibly even earlier record was listed by Luppold et al. (1994) from Müssenberg (Bed 1). Eickhoff (1973) noted at Oberrödinghausen a joint entry of Ps. triangulus inaequalis and Ps. triangulus triangulus in his Sample 3, correlated with Sample 5 of Voges (1959), which is in the S. (S.) duplicata Zone. More precision of the Ps. triangulus ssp.-siphonodellid correlation is required.
Ammonoid biostratigraphy (Fig. 1) The DCB ammonoid biostratigraphy has been reviewed by Becker et al. (2016a), who emphasised parallel zonations based on successive members of the Prionoceratidae, Kosmoclymeniidae, other clymeniid groups, and, in the lower Tournaisian, early prolecanitids. A different approach was followed by Klein and Korn (2016), who used three quantitative methods (unitary associations, ranking and scaling, constrained optimisation) to correlate Rhenish records (reference sections Müssenberg 1 for the upper/uppermost Famennian, Oberrödinghausen Railway Cut for the lower Tournaisian) and to establish composite ranges. These, however, differ from each other and the correlation obviously did not consider other stratigraphic markers, such as conodonts and metabentonites. The Rhenish Slate Mountains yielded on a global scale the most detailed DCB ammonoid record. Several ammonoid subzones are based on the detailed classical studies by Schindewolf (1937) and Vöhringer (1960). Subsequent studies and revisions added many precise data and taxa (e.g. Korn , b, 1984Korn , 1988aKorn , 1988bKorn , 1991Korn , b, 1993Becker 1988Clausen et al. 1989a, b;Korn and Luppold 1987;Luppold et al. 1994;Korn et al. 1994b;Korn and Weyer 2003;Korn and Vöhringer 2004;Ebbighausen and Korn 2007;Becker et al. 2016a) but in general supported the originally established faunal sequence. The current knowledge of Rhenish DCB ammonoid zones/subzones, with all units (apart from the Post. evoluta Zone) named after the FAD of its index species, is from base to top (for reference section literature, see section descriptions): Linguaclymenia similis Zone , UD VI-A 1 ) Synonyms: Muessenbiaergia sublaevis Zone . Discussion: Schindewolf (1937) proposed to use the entry of Kalloclymenia (s.l., at his time including various sphenoclymeniids and Finiclymenia) to define the base of the Wocklum Beds and a Kalloclymenia-Wocklumeria-Stufe. However, in his type section in the Oberrödinghausen Railway Cut, bed-by-bed collecting started with Bed 23, from which he recorded "Oxyclymenia bisulcata". At his time, such an identification referred to kosmoclymeniids with marked ventrolateral furrows, which included, among others, forms now assigned to Linguaclymenia, especially L. similis (see synonymy list in Korn and Price 1987). Therefore, it can be assumed that Schindewolf's section began locally with the FOD of Linguaclymenia. The first "Kalloclymenia", in fact "Sph." cf. brevispina, commenced one bed higher, which led to the term K. subarmata-K. brevispina Zone. Korn and Luppold (1987) subdivided the zone and defined their "Lower subarmata Zone" by the entry of "Sph." brevispina, stimulated by new records of that species from Dasberg and Effenberg. However, in the subsequent review of "Sph." brevispina by Price and Korn (1989), the Dasberg specimens were re-assigned to a new species, "Sph." erinacea (see also . This showed that several "sphenoclymeniids" enter in the basal Wocklum Beds. The phrase "Sphenoclymenia" refers to the distinction of moderately large forms with narrowly rounded venter from the poorly known, giant-sized, oxyconic-type species of the genus, Sph. maxima. At Müssenberg Klein and Korn 2016), M. sublaevis and L. similis enter jointly in Bed 90. This was obviously the base for a re-definition of the Wocklum-Stufe by the FAD of M. sublaevis , although "Sph." brevispina enters at Müssenberg five beds higher. Because kosmoclymeniids are more common than "sphenoclymeniids", we accept Korn's slight extension of the Wocklumeria-Stufe but give preference to L. similis as the main index species. It is very distinctive and more easily recognisable than M. sublaevis in poorly preserved faunas without shell, even as small fragment. In addition, its zone can be traced widely, for example to Morocco . A joint FAD of L. similis and M. sublaevis was also shown at Dasberg . Subdivision:  re-habilitated Kalloclymenia (s.str.) as a marker genus for the lower Wocklumian. It enters at Oberrödinghausen : K. cf. uhligi), Müssenberg (Bed 78: K. subarmata), and Hasselbachtal (Bed 2: K. pessoides) well above the base of the L. similis Zone (compare ranges in Klein and Korn 2016). However, Korn and Luppold (1987) noted a Kalloclymenia sp. from the second bed of the Wocklum Limestone at Dasberg.
The complete evidence is a mosaic from these different sections. It suggests that the base of the Bi. ultimus ultimus Zone, if defined by either of the three main markers, lies close to the base of the L. similis Zone. Bispathodus ultimus ultimus, the proposed index fossil for the definition of a formal uppermost Famennian substage, however, does not display consistent FODs in individual sections and partly very long delays. Its oldest regional occurrence post-dates the oldest "Sph." brevispina. This suggests that Bi. ultimus ultimus was ecologically sensitive in pelagic facies.

Muessenbiaergia bisulcata Zone (Korn 2000; UD VI-A 2 )
Synonyms: Upper subarmata Zone, lower part . Discussion: Already  and Luppold et al. (1994) recognised the level of Eff. lens and Eff. minutula as an upper subdivision of their "Late K. subarmata Zone" but the zone was not formally named until . It can be followed to southern Morocco . Conodont correlation: Based on the Müssenberg section , the base of the Eff. lens Zone (Bed 38) lies well above the FAD (Bed 55) of S. (Eosiphonodella). This is supported by less detailed data from Oese .
This summary suggests that Ac. (Str.) carinatum, Ac. Zone. However, this is based on combined evidence, not on a locality with a faunal sequence. Therefore, we regard the current knowledge as too premature to propose subzones. The type level of the poorly known Ac. (St.) infracarbonicum (Paeckelmann, 1913)  Gattendorfia subinvoluta Zone and Subzone (Vöhringer 1960, LC I-A 2 ) Synonym: Acutimitoceras (Acutimitoceras) acutum Zone (Vöhringer 1960). Discussion: As noted in the taxonomic appendix, there are morphological differences between the types of G. subinvoluta from Franconia and assumed Rhenish representatives of the species. This may reflect the biogeographic separation by a narrow oceanic system.

Paragattendorfia patens Zone (LC I-D)
Reference sections: Oberrödinghausen Railway Cut , Oese . Other marker species: Eocanites supradevonicus (Oberrödinghausen, Bed 2; Oese, Bed S = Korn-Bed 36, Kullmann in Becker et al. 1993). Korn and Weyer (2003), however, noted an Eoc. cf. supradevonicus already lower at Hasselbachtal (Bed 51H), in association with the last local Paprothites (top LC I-B). Following   fig. 2), Paralytoceras crispum enters at Oberrödinghausen at the base of the patens Zone. Subdivision: The entries of Eoc. planus and Pseud. serratus in Bed 1 of Oberrödinghausen can possibly be used to define an upper subzone. However, since the type level is relatively poor in fauna (Vöhringer 1960), additional data are needed. Korn (1988a)  Massif as examples/reference, the new scheme will add precision to global correlation and provide distinctive, unequivocal terms to be used during the ongoing GSSP search. Refined information from other regions may lead to the recognition of further event stratigraphical units. The subdivision of the Hangenberg Crisis Interval in the Rhenish Massif (from base to top) is as follows: Crisis Prelude (initial regressive interval) Part I. Drewer Sandstone with Post. evoluta or increasing condensation and macrofossil accumulation in the upper Wocklum Limestone, top W. denckmanni Subzone (UD VI-D 1 ); supposed equivalents in the neritic facies of the Velbert Anticline (see Paul 1937aPaul , 1939a fig. 6C, with simplified lithology of limestone nodules and nodular limestone in white, shale and siltstone in grey, black shale = Bed 115N, and dark-grey marl = Bed 85H). 1 = chronostratigraphy, 2 = ammonoid zonal key after , 3 = conodont zonation, 4 = bed numbers, 5 = lithology, TOC = total organic carbon.

Stable isotope stratigraphy
In the Rhenish Massif, DCB carbon and oxygen stable isotope analyses were conducted bed-by-bed at Hasselbachtal (Figs. 4,5), Oberrödinghausen (Fig. 5), and Oese Kaiser et al. 2006). Carbon isotopes of micrites (δ 13 C carb ) and of sedimentary organic matter (δ 13 C org ) were used to reconstruct variations in the carbon isotope composition of the oceanic dissolved inorganic carbon pool. During marine anoxia, global black shale deposition (HBS), and enhanced organic carbon burial, a pronounced positive carbon isotope excursion (CIE) could be recorded at Hasselbachtal (Fig. 4). The CIE is well correlated with a CIE measured in black shales at Kronhofgraben (Carnic Alps), in continuous pelagic limestone successions at Grüne Schneid (Carnic Alps), as well as with positive peaks measured in several other distant regions worldwide (see review in Kaiser et al. 2015). Thus, carbon isotopes can be used as a chemostratigraphic tool to record the level of the anoxic Lower Crisis Interval (HBS) even when there are locally no black shales. Conodont apatite is used for the analyses of δ 18 O phosph because it has a high preservation potential of the primary isotopic signature and is considered to be unaffected by diagenesis. The oxygen isotopic composition of biogenic phosphate is dependent on temperature and composition of ambient seawater, and can be therefore used as a proxy for estimating sea-surface temperatures and salinity changes. The high abundance and high morphological diversity of conodonts provides a detailed bio-and chemostratigraphic resolution and the correlation of sections from different palaeogeographical realms. In this respect, the oxygen isotope values measured from Hasselbachtal, Oese, and Oberrödinghausen ( Fig. 5) are similar to values from other sections in Spain (Pyrenees) and France (Montagne Noire). Slightly decreasing values from the Bi. ultimus ultimus Zone to the lower Tournaisian indicate a minor warming during the corresponding time span, if there was no influence of palaeosalinity, and without details for the intervening Hangenberg Crisis Interval. In the Rhenish Massif, no conodonts were available from the ckI. Records of a cooling episode, probably time-equivalent to the regressive Middle Crisis Interval, were observed in equivalent successions of the Carnic Alps ).

Sequence stratigraphy
The successions of the Rhenish Massif have been used as a reference for DCB sequence stratigraphy ( Van Steenwinkel 1993;Bless et al. 1993;Kaiser et al. , 2015. A gradual shallowing upwards of the depositional environments is evident in the upper part of the Wocklum Limestone by increasing condensation (reduced sedimentation rates), a consequently strongly increasing fossil content, and by the return of shallow pelagic taxa, such as large-eyed trilobites or tabulate corals ). This highstand episode was followed by the local "Drewer Sandstone" (Crisis Prelude I, Fig. 1), which can be regarded as a lowstand deposit and its base represents a (para)sequence boundary. In the neritic realm of the eastern Velbert Anticline (Fig. 6), there are correlative pre-Crisis oolites and sandy beds, changing laterally (Wuppertal region) into red shales, marls, and limestones (Am Haken Brickwork Quarry). In several Sauerland sections (Dasberg, Reigern, Kattensiepen), the upper part of the  "1" to "3" = correlation levels sensu ; ammonoid zonal key (e.g. VI-B, I-C) as explained in the text, ckI costatus-kockeli Interregnum, kock. kockeli, brans.
bransoni, dupl. duplicata, sdbg. sandbergi, lob. lobata M1, cren. crenulata, m. T. middle Tournaisian Wocklumian falls in a time of current-induced non-deposition (Fig. 7). The shallowing at the "Drewer Sandstone" level can be recognised in many parts of the world and represents a global pre-Crisis sea-level drop (Kaiser et al. , 2015, possibly related to an initial glacial cooling. The Hangenberg Black Shale and its less anoxic, grey to greenish equivalents in western sections (Cromford, Velbert 4 Well, B224 near Aprath; Fig. 6) represent a sudden transgressive episode and phase of maximum flooding, defining the Lower Crisis Interval. The pyrite-enriched, often sharp base of the black shales suggests an anoxic marine flooding surface at the base of a TST, which is the level of the main Hangenberg Extinction, at least in the pelagic realm. The transgressive nature of the HBS is underlined by sections, where greenish, dark-grey, or black (hemi)pelagic shales, often with Post. cf. evoluta, overly shallow-water strata, such as neritic limestones (Cromford, Velbert 4 Well, Fig. 6), oncoidal siltstones (B224 near Aprath, Fig. 6), or sandstones (Vingerhoets 93 Well, Fig. 7). Similar patterns can be seen on other continents (Kaiser et al. 2015;Zhang et al. 2019).
The gradual change from the HBS to greenish and silty shale marks the slowing of sea-level rise to beginning sealevel fall and regression. The Hangenberg Shale is a pelagic highstand deposit (FSST) still low in the LN miospore Zone (Middle Crisis Interval I).
A subsequent main lowstand phase (LST) is represented by the locally increasing influx of coarser clastic sediments, the Hangenberg Sandstone, and its equivalents, such as siltstones (Drewer, Borkewehr) or thick, re-sedimented oolitic conglomerates and siliciclastics of the Seiler area (Fig. 7). A sequence boundary is placed at the base of the sandstones, conglomerates, or reworked oolites. The latter represents basin floor fans and incised valley fills above marine unconformities (e.g. Van Steenwinkel 1993). Herbig (2016) used the Hangenberg Sandstone base to define his Sequence 1. Both in the western realm (Aachen region: Binsfeldhammer, Velbert Anticline: Langenhorst; Fig. 6) and in the deeper facies of the Sauerland (Müssenberg, Kattensiepen, Drewer; Fig. 7), the sea-level fall caused gaps due to erosion and/or non-deposition. The partly much thicker HSS equivalent clastic deposits of other regions and continents correspond with a maximum lowstand (LST) in the higher LN miospore Zone and upper ckI (Middle Crisis Interval II).
In the eastern Rhenish Massif, the Upper Crisis Interval (DCB interval) is characterised by a return to carbonate sedimentation and condensed, fully pelagic conditions. This transgressive (TST) interval is characterised by the polyphase Stockum Limestone deposits, including some turbiditic limestones (Hasselbachtal, Oese, Apricke; Fig. 7), or by partially correlative, pyritic, or organic-rich Stockum Level Black Shales (SLBS, Am Haken Brickwork Quarry, Hasselbachtal; Figs. 6, 7), sometimes with embedded, fossiliferous limestone nodules (B224 near Aprath, Drewer). For the precise DCB positioning and correlation, it is important to realise that the first limestone level above the HS/HSS is locally of variable age (Fig. 7), from the top ckI with Pr. collinsoni (Drewer, ?Eulenspiegel; Upper Crisis Interval I), to the Pr. kockeli  (Kaiser et al. 2015). The Hangenberg Limestone represents a highstand interval (HST), in which the fossil content becomes gradually sparser, changing eventually into a shaly interval. The sharp base of the anoxic to euxinic, basal middle Tournaisian Lower Alum Shale (LAS) represents the very sudden onset of a second Lower Carboniferous TST (base of Sequence 2 of Herbig 2016). It is equally distinctive in the western neritic realm, where it is expressed by the less anoxic and party fossiliferous Pont d'Arcole Formation that drowned the Hastière Formation carbonate platform (Fig. 6).
Trace element geochemistry, magnetic susceptibility, and cyclostratigraphy Trace element geochemistry and gamma ray spectroscopy (CGR values) are important tools to reconstruct changes of detrital discharge, redox conditions (U/Th), and palaeoproductivity around the DCB. Magnetic susceptibility (MS) logging provides insights into changes of sedimen-  tation, especially in relation to sea-level change. These methods, therefore, complement litho-and sequence stratigraphy. In addition, they are most useful to delineate cyclostratigraphic patterns, which are strongly expressed in the Wocklum and Hangenberg Limestones. For refined cyclostratigraphic correlations, it will be most important to tie each cycle to precise bed numberings and the biostratigraphical scales (see Korn and Weyer 2003). Kumpan et al. (2015), Becker et al. (2016b), and Hartenfels et al. (2017a, b) provided relevant data for four Rhenish DCB sections: Oese, Oberrödinghausen, Drewer, and Borkewehr. They support the general interpretations of sedimentary environment, microfacies, palaeoecology, sea-level changes, and event patterns as summarised by Kaiser et al. (2015).

Aachen region and Niederrhein
In terms of sedimentation and lithostratigraphy, the DCB transition of the Aachen region continues the shallow-water setting of the eastern Ardennes. The Belgian formation terminology (e.g. Bultynck and Dejonghe 2002;Poty et al. 2002) can be largely adopted. The term Dolhain Formation, used for equivalents of the Etroeungt Formation in the Vesdre Nappe of Belgium (e.g. Bultynck and Dejonghe 2002), has not yet been accepted on the German side (e.g. Amler and Herbig 2006;Kasig and Reissner 2008). Currently, the only moderately good DCB outcrop is at Binsfeldhammer in the Burgholz Syncline.

Summary of succession:
Pont d'Arcole Formation (previously "peracuta Shale" or "Zwischenschiefer"): Unfossiliferous dark shales, with the major deepening of the Lower Alum Shale Event at the base; middle Tournaisian. Hastière Formation, Binsfeldhammer Member: Ca. 3.5 m of cross-bedded dolomites with rare syringoporid corals above the last level with stromatoporoids; with Pr. kockeli in the lower part and S. (S.) duplicata in the higher part in the lateral Neu-Moresnet section (Reissner 1990). Age: Pr. kockeli Zone (Upper Crisis Interval) to top of lower Tournaisian. Our new sampling at Binsfeldhammer did not produce any conodonts.
Lower to middle parts: Calcareous shales, thin-bedded limestones and thin-to thick-bedded dolomites with brachiopods, corals, stromatoporids, and foraminifers. Quasiendothyra kobeitusana, the index species of the DFZ 7 and for the uppermost Famennian, begins ca. 9 m below the top of the formation. Discussion: Dolomitisation and poor outcrop conditions prevent an easy recognition of the unconformity. However, there is locally no evidence for sediments of the Lower/Middle Crisis Interval. The Hangenberg Regression may have led to erosion and/or non-deposition. The presence of an unconformity was supported by the analysis and correlation of coral faunas (Weyer 2000). Mottequin and Poty (2014) mentioned from their adjacent Stolberg Section a 2m thick sandstonesiltstone unit overlying Etroeungt (= Dolhain Fm.) dolomites. This unit, which is unknown at Binsfeldhammer, was interpreted by them as HSS equivalent.
Velbert Anticline DCB successions along the northern margin of the Velbert Anticline are characterised by a lateral facies transition (e.g. Franke et al. 1975;Herbig et al. 2001). The upper part of the thick Velbert Formation is characterised in the Ratingen area by the increasing intercalation of limestones with neritic fauna, such as rugose corals, brachiopods, phacopids, stromatoporids, ostracods, and foraminifers. Following Drevermann (1902), Bärtling and Paeckelmann (1928) and Paul (1939a), the term Etroeungt Member is revived for this succession (the former "Etroeungt-Schichten"). It is not wise to neglect this interval of reduced clastic input within an undivided Velbert Formation due to its importance for correlation towards the Ardennes. Because it is unsuitable as a mapping unit, it cannot be separated regionally as a full formation, unlike as in the Ardennes. Future facies and faunal studies may support a lithostratigraphic differentiation between the type Etroeungt Formation of the southern Dinant Syncline, the dolomitised neritic facies settings of the Vesdre-Aachen region (proposed Dolhain Formation), and the Velbert Anticline. There, the term "Angertal Schichten" was established as a lower subdivision of his Etroeungt by Paul (1939a) on faunal, not on lithological, grounds, but it is principially available as a member name if neritic workers agree on a lithostratigraphical definition and type section. As outlined by Amler and Herbig (2006), the uppermost Famennian limestones grade east of Heiligenhaus into a siliciclastic facies with an important influx of oolithic material, especially in the lower Tournaisian (Laupen Member). To the NNW of the Velbert Anticline, DCB beds were once also accessible at two small distinctive saddles of the Lintorf region (sheet 4607 Kettwig). For example, Paul (1938a) suggested the presence of Postclymenia in the Diepenbrock area.
Bed 16, 8.7 m light-to dark-grey, medium-to thickbedded, solid crinoidal limestone with stromatoporids (type level of Clathrodictyon ratingense Paul, 1937a, pectinid bivalves, brachiopods, solitary rugose corals ("Palaeosmilia" aquis-granensis), syringoporids, foraminifers (Quasiendothyra kobeitusana), pre-Crisis Interval, uppermost Famennian. Discussion: It is very unfortunate that the critic interval, from very fossiliferous limestones of the assumed Crisis Prelude to the shale with Postclymenia and basal Hastière Formation, is currently not accessible. The Cromford section provided unique opportunities for the correlation and intergradation of Rhenish pelagic and neritic successions and for the understanding of DCB extinction patterns in the two shelf realms. As evident from the comments by Winkler-Prins and , the Cromford brachiopods require revision, which is true for most other faunal elements from the locality. Paul (1939a) mentioned shales with Post. cf. evoluta from numerous other adjacent localities on sheet Kettwig: valley below Sondert, debris of the Thalburg Mine, Ruthen, Giesenhaus area, road slope between Roßdelle and Müllerbaum, valley end at Scharpenhaus, Herberg, and Steinloch Mine. It is remarkable that the pure or calcareous Postclymenia shales, the supposed shallow-water Hangenberg Black Shale equivalents, yielded also a rich, oxic neritic fauna, consisting of diverse brachiopods (productids, orthids, spiriferids, terebratulids), bivalves (pectinids, nuculids, Grammatodon), fenestellid bryozoans, and gastropods. It is especially regrettable that a directly underlying, 10 cm thick bluish-grey, crystalline limestone with numerous Postclymenia from Scharpenhaus (Paul 1939a, p. 676, Bed 2) cannot be sampled any more for conodonts, microfacies, or geochemistry.
Discussion: It is intriguing that stromatoporids locally do not range to the top of the Velbert Formation. The presence of "Strunian"-type fossils (corals, quasiendothyrids) in the lower part of the Hastière Formation resembles the Cromford succession. It questions the precise extinction level of shallowwater faunas in relation to the pelagic extinction at the base of the Hangenberg Black Shale. Therefore, Denayer et al. (2019) suggested that neritic extinctions occurred on the Ardenne Shelf during the maximum regression ("Hangenberg Sandstone Event"). However, since Pr. kockeli enters already in the basal marker limestone of the Hastière Formation (e.g. Bouckaert and Groessens 1976), the survival of various neritic taxa lasted obviously well into the transgressive Upper Crisis Interval. This explains why at Klein-Steinkothen no major regression can be recognised within the exposed Hastière Formation above the level of the last quasiendothyrids and "Strunian"-type corals. In the absence of age-diagnostic conodonts or brachiopods, the timing and patterns of neritic extinction are locally difficult to resolve.

Hefel
Location: Disused, partly pond-filled old quarry N of Velbert (Fig. 9), a natural reserve, at the small settlement Hefel, sheet 4608 Velbert, r 2 573 340, h 5 691 780. Literature: Bärtling and Paeckelmann (1928), Gallwitz (1932), Paul (1937a), Franke et al. (1975), Paproth et al. (1976), Paproth and Streel (1982), Herbig et al. (2001), Amler and Herbig (2006). Succession: The 58 m thick succession ranges from the Etroeungt Member to the Richrath Formation. The DCB part is strongly condensed and incomplete. Paproth et al. (1976) illustrated two specimens of Post. cf. evoluta (as Cymaclymenia euryomphala) from a calcareous silt/sandstone at the base (Bed 1), which may correlate with the "Drewer Sandstone" (Crisis Prelude). This tentative interpretation is based on the fact that the Postclymenia level was overlain by probable Hangenberg Shale equivalents (Beds 4-6), followed by thin oolites (Beds 4-6, Laupen Member of Hastière Formation). The section requires new investigations but it is doubtful that the Postclymenia level is still accessible. For example, Amler and Herbig (2006) observed that the oolithic basal Laupen Member eroded into the top limestone of the "Strunian". Velbert Formation, Etroeungt Member: 30.5 m bioclastic, tempestitic, marly limestone with crinoids, bryozoans, brachiopods, and ostracods, but lacking calcareous algae and age-diagnostic foraminifers Discussion: The special importance lies in the local recognition of a HBS to HSS sequence, which provides a correlation with the Sauerland-type succession. Unfortunately, there is so far no record of biostratigraphical markers.

Wuppertal area (western Remscheid-Altena Anticline)
B224 outcrops NE of Aprath (Fig. 6) Location: Temporary outcrops were created during excavations and trenching for the B224 road, which crosses the western part of the narrow Herzkampe Syncline ESE of Wülfrath and NW of Wuppertal-Elberfeld (Fig. 9). A multidisciplinary approach led to a voluminous monograph on the Upper Devonian to Lower Carboniferous (Thomas 1992). The uppermost Devonian und lower Tournaisian were developed as calcareous and micaceous shales and siltstones, which differ from the typical sequence of the Velbert Anticline. The succession, which has no current outcrop, can be summarised as follows (see Korn and Thomas 1988;Thomas 1992;Korn 1992b Beds 05/21-29, 3.5 m calcareous siltstones with rare oncoids. Bed 05/20, 70 cm calcareous siltstones with oncoidal rud-/ packstones and neritic fauna (brachiopods, gastropods, small solitary Rugosa), ?Crisis Prelude.
The most westernly occurrences of Stockum Limestone, interrupting a succession with few pelagic fossils, underline its transgressive nature. The lower level of large septaria and ammonoids at the B224 may express locally the transgressive Hangenberg Black Shale interval. It is unfortunate that there are no conodont and miospore data and that there are no available outcrops.
1 m grey shale with thin sandstones, probably Crisis Prelude I. 20 cm grey and reddish limestone with Richterina and Drevermannia, upper hemisphaerica Zone, pre-Crisis Interval.
-unconformity (no record of the lower Tournaisian VI Zone or Hangenberg Limestone equivalents) -(proposed) Brahm Formation (previously "Obere Cypridinenschiefer"), Member 4 (Unit 1 of Zimmerle et al. 1980): Ca. 9 m alternating grey, silty shales, and fine-to middle-grained calciturbidites with poor macrofauna   (1926) recorded Weyerites wocklumeriae, a typical uppermost Famennian phacopid, from a dark-grey limestone. Paeckelmann (1923, p. 284) as well as Fuchs and Paeckelmann (1928, p. 38) noted the presence of Postclymenia at Riescheid. This indicates that time equivalents of the Hangenberg Black Shale may be distinguishable in the lower LN Zone interval. The DCB unconformity disqualifies Riescheid as a possible GSSP Auxiliary GSSP, showing occurrences of index conodonts and ammonoids, the position of marker bentonites (ω 1-3 and a 1-3 ), and their geochronological ages, miospore zones, and changes of carbon isotopes (for references see text, for lithology legend, see Fig. 13) section but its importance lies in the miospore record, which provides marine-terrestrial correlations.

Northern Sauerland (eastern Remscheid-Altena Anticline)
In the northern Sauerland, between Hagen in the west and Warstein in the east, the "standard" DCB succession (see summary in Becker et al. 2016a) is developed. A major channel fill, the Seiler Conglomerate, forms a short W-E interruption in the Iserlohn area (e.g. Gallwitz 1928;Ziegler 1970;Paproth 1986). The widely used traditional lithological terms do not conform to modern lithostratigraphical standards. The Wocklum Beds or Wocklum Limestone have been based on its distinctive faunas (Wocklumian, UD VI; Schindewolf 1937), not on a sedimentary change that would enable a clear separation from the underlying Dasberg Beds (now Dasberg Formation). Attempts to identify a subordinate marker in the best and continuous sections (Oese, Oberrödinghausen, Effenberg), in order to justify at least a member status, failed so far. In the type section near Wocklum (see below), there is a reddish unit with large kalloclymeniids above a faulted zone. At Oese, a possibly correlative, peculiar interval of red shales and red nodules occurs above the oldest Wocklumian ammonoids. However, we did not trace it in the other sections. Consequently, we continue the use of the term Wocklum Limestone in a traditional, biostratigraphically defined form. The whole interval from the base of the Hangenberg Black Shale to the base of the Lower Alum Shale is assigned to the Hangenberg Formation, which is subdivided into informal members.
Bed 109N with the last Pa. gracilis gonioclymeniae (Stoppel in Becker et al. 1984; top of former Middle costatus Zone).
Bed 101N with a member of the Cyrto. procera Group (new record).
Bed 8S with Br. suprema (delayed FOD in the upper Bi. ultimus ultimus Zone).
Bed 0S with a member of the giant-sized "K." pachydisca Group.
3 m below Bed 0S, "Mim." liratum (L. similis Zone, basal UD VI-A 1 ). Discussion: Since the northern and critical DCB part of the outcrop is currently covered, it cannot be considered in the new GSSP search. Attempts to trench a new exposure on lateral public ground was unsuccessful. In addition, Bed 85H, the basal part of the transgressive Crisis Interval III, was not calcareous enough to produce conodont faunas that could shed light on the Pr. collinsoni-Pr. kockeli transition. The thin metabentonites enable a high-precision correlation along the northern limb of the Remscheid-Altena Anticline (Hasselbachtal to Oese and Apricke, Korn and Weyer 2003). So far, only two of them have been dated geochronologically (Trapp et al. 2004: 360.5 ± 0.8 ma for Bed 79H, 360.2 ± 0.7 ma for Bed 70H).
Trench 2, 54.6-73.5 m, two packages of micaceous sandstone, in the lower part up to 15 cm thick beds, interrupted by a ca. 4 m thick silty shale interval.
Trench 2, 266.0-266.4 m, coarse, light-grey limestone conglomerate with calcareous matrix, Pa. gracilis gonioclymeniae, Bi. costatus costatus, and many reworked lower/middle Famennian conodonts (Sample 57, basal Bi. ultimus ultimus Zone). Discussion: Conodont faunas from the post-Crisis Hangenberg Limestone of the Seiler are most important to understand the Rhenish lower Tournaisian siphonodellid succession. Therefore, it is very unfortunate that no re-sampling is currently possible. Another interesting feature is the brief report of a detrital black limestone at the base of the Crisis Interval, which, unfortunately, was not studied in detail. It may  Fig. 1; for references to the numerous records, see text,* = reworked; for lithology legend, see Fig. 13 represent one of the globally rare limestones deposited during the main anoxic pulse.
at the base could be used for a lithostratigraphic distinction of the overlying Wocklum Limestone but has not yet been correlated with the Oberrödinghausen section. The lower reddish interval  may correlate with a distinctive reddish unit at Wocklum that yielded large-sized Kalloclymenia.
Bed 8 with the P. paradoxa Zone (UD VI-C). Discussion: Beds 1H-6H (39-36) should be re-sampled for conodonts in order to find a Protognathodus succession. Korn and Weyer (2003) proposed to correlate the turbiditic Bed 1H with the turbiditic Bed 84H at Hasselbachtachtal. If this is correct, Beds 2H-6H at Apricke correlate with the Upper Stockum Limestone interval  at Hasselbachtal, which is supported by the overlying thin metabentonite a1 in both sections. However, it is also possible that the thin Hasselbachtal turbidite (Bed 84H) correlates with the upper thin turbidites at Apricke (Beds 4H/5H).
A corresponding, adjacent section that is now overgrown was exposed for a long time along the main road through the Hönne Valley Korn and Weyer 2003; both with section logs). The top Wocklum Limestone was very fossiliferous (with abundant W. denckmanni, F. wocklumensis, Cyrtoclymenia, Gl. glaucopis, and other ammonoids, RTB collection). Korn and Weyer (2003) mention an unpublished collection of ca. 3.000 ammonoid specimens from the Wocklum Limestone by a Tübingen Ph.D. student, and of further 500 specimens collected by D. Korn. This material includes probably further interesting records. Despite the only 200 m distance to the railway cut, the Hangenberg Shale is thicker and contains massive Hangenberg Sandstone beds. Two research boreholes, Oberrödinghausen 1 and 2, were drilled in 1979 and studied for conodonts (zonal identifications by W. Ziegler) and palynology (Higgs and Streel in Becker et al. 1993). Discussion: The local DCB unconformity excludes the section from a re-consideration as a GSSP candidate, despite its enormous significance for ammonoid-conodont stratigraphy/ correlation at levels below and above the Hangenberg Crisis Interval. This situation could have changed if the Hangenberg Extinction had been re-considered as the future DCB level, as proposed first by Paul (1937b, footnote on p. 748), and then by Walliser (1984). However, the majority of the DCB Task Group voted against this option in 2016.
The re-sampling of the basal Hangenberg Limestone by Sacher (2016: Bed 227)   . It shows that there is a thin layer of Upper Stockum Limestone preserved and that the hiatus at the base of the limestone successions corresponds to the DCB regression known from Hasselbachtal. Luppold et al. (1994) reported a different fauna with Pr. kockeli, S. (Eo.) sulcata s.l. and different reworked Famennian conodonts from their corresponding Bed 6A. This faunal difference points to a complex mixture of reworking and condensation. It needs to be stressed that there is no record of any species of the Gattendorfia Fauna in Beds 6b = 6A = 227.
Hangenberg (Fig. 7) Location: Trenches from 1973 at the Hangenberg near Deinstrop, northern Sauerland (Fig. 9), topographic sheet 4613 Balve, r 3 423 900/h 5 695 750. Literature: Schmidt (1924a, b), Paeckelmann (1924), Paeckelmann and Kühne (1938), Stoppel 1977, Luppold et al. (1994. Succession: Although this was the original type area for the Hangenberg Beds (now Hangenberg Formation) of Schmidt (1924a, b), there is hardly any outcrop. A trench dug in 1973 had to be filled quickly; it was measured by Leuteritz and Schäfer and described by Stoppel 1977 provided details for natural outcrops in the vicinity. Both sources together give the following succession: Kahlenberg Formation (Lower Alum Shale): Unfossiliferous black shale, Lower Alum Shale Event Interval, middle Tournaisian. Hangenberg Limestone: 1.3 m grey limestone, compact in the lower part, more nodular in the upper part, rich in conodonts in the first thick bed (Samples III-X) but still without zonation for the main and upper parts; with protognathodids ( Fig. 2a-d).
A few cm of solid limestone (Sample II)  Bed of bluish-grey, argillaceous limestone without any fauna; local Middle Crisis Interval Ib.
Ca. 6.5 m platy, olive-grey to dark-grey shale, local Middle Crisis Interval Ia.
Bed E, 19 cm, three layers of nodular limestone and intervening shale with Mim. trizonatum.
Bed F, 22 cm, seven layers of nodular limestone with Eff. lens (base Eff. lens Zone, UD VI-B).
Bed I, 22 cm, six levels of limestone nodules and nodular limestone with M. bisulcata and M. parundulata (base M. bisulcata Zone, UD VI-A 2 ).

Reigern Quarry, Hachen
Location: Mostly filled and overgrown old quarry in the forest ca. 500 m N of Hachen, northern Sauerland ( Fig. 9), topographic sheet 4613 Balve, r 3 429 770/h 5 695 230 (GPS 51°23′ 15.77″, N 7°59′ 25.20″). Literature: See compilation in the recent review by Söte et al. (2017). Subsequently, a Sphenoclymenia brevispina from Reigern was figured by Korn and Price (2019). Succession: The Reigern Quarry is a very important locality for ammonoids from the lower part of the Wocklum Limestone (Korn , 1988bPrice 1982;Korn and Price 1987;Price and Korn 1989). The recent conodont study by Söte et al. (2017) proved the presence of the Bi. ultimus ultimus (Beds 1-23) and S. (Eo.) praesulcata s.l. Zones . The DCB is marked by a long-lasting major unconformity, with non-deposition starting before the base of the P. paprothae Subzone (UD VI-C 1 ) and lasting until the Viséan. Fig. 13 Summary section log for the DCB transition at Stockum (redrawn from the section log of Clausen et al. 1994, fig. 6) showing the occurrences of marker conodonts, ammonoids, and miospore zones. For abbreviated zones and lithological units see Fig. 1 Beds 120-157 (lower/middle part of Unit Alb-C), ca. 9.5 m greenish-grey, silty, micaceous shale alternating with thin, partly laminated or calcareous, micaceous silt and sandstones, with a solid sideritic level (Bed 153) and a weathered, darkgrey, argillaceous, unfossiliferous limestone (Bed 152)   , Lower Crisis Interval. Basinal Shale equivalents of Wocklum Limestone (no modern formation name): Beds 163-187 (lower/middle Unit Alb-B), 7.2 m greenishgrey to brownish, silty shales with thin, dark-grey limestones in the middle (Bed 172) and upper parts  The initial major significance of the Stockum Limestone lenses was the discovery of a distinctive conodont and ammonoid fauna that differs strictly from the pre-Crisis (typical topmost Devonian) Wocklumeria and post-Crisis (typical basal Carboniferous) Gattendorfia faunas. Since a similar fauna was lacking at Oberrödinghausen, it became clear that the stratotype includes a significant gap right at the boundary, which disqualified it and led to the search for a new DCB GSSP. Subsequently, a succession of Protognathodus faunas (meischneri/collinsoni-kockeli-kuehni) was established, which shows that the larger Stockum Limestone interval comprises in various localities up to three different levels (here: Upper Crisis Interval I-III). At Stockum, "older" (without Pr. kuehni) and "younger" (with Pr. kuehni) Protognathodus faunas were distinguished since ; they characterise the Lower and Upper Stockum Limestone sensu Becker et al. (2016a). The resampling by Clausen et al. (1994) of the main goniatite limestone of Schmidt (1924a, b), Bed 103, did not reproduce the  and Groos-Uffenorde and Uffenorde (1974) records of Pr. kuehni. This led to the assumption that the kuehni level was the slightly younger, thin Bed 100 of Clausen et al. (1994), which is also rich in goniatites, but mostly only in juveniles. The presence of Po. marginvolutus and of pseudopolygnathids in Bed 103 Stoppel in Clausen et al. 1994, Streel and is not typical for conodont faunas of the kockeli Zone of other Rhenish localities.
E of Weninghausen (Fig. 7) Location: Former slopes along a track branching off to the N from the road from Linnepe to Wenighausen, sheet 4614 Arnsberg, ca. r 3 435 120 /h5 687 125 (Fig. 9). Literature: Gallwitz (1928), Kühne (1938). Succession: The shelf basin facies between the Stockum area and the seamount settings of the drowned Warstein, Belecke, and Brilon reefs towards the E/NE is mostly poor in  , here updated from Becker et al. (2016b). For abbreviated zones and lithological units see Fig. 1; for references to the numerous records see text, * = reworked; for lithology legend see Fig. 13 culminidirectus (upper part of Pr. Bed 94, 40 cm organic-rich, black, fissile alum shale, poor in macrofauna but with rare Post. cf. evoluta, base of Post. evoluta Zone (UD VI-E) and lower ckI (no conodonts so far), maximum flooding (low Zr/Al ratio), anoxic (high U/Th ratios), Lower Crisis Interval II.
In the 1978 section of , the top of the Wocklum Limestone is followed by a bed (Sample 7) with mixed, dominant pre-crisis conodonts and a single Pr. kockeli. The next bed (Sample 6) had already S. (S.) bransoni. Therefore, the critical DCB interval was missing by extreme condensation. The situation was not much better in their section sampled in 1979. Discussion: Drewer is the only Rhenish section, where the shallowing upwards in the upper Wocklum Limestone led to a siltstone intercalation, the "Drewer Sandstone", which is a good marker for the base of the Crisis Prelude. The Middle Crisis Interval is locally very condensed, possibly due to an unconformity at the base of the silty equivalent of the Hangenberg Sandstone. Therefore, there is locally no Hangenberg Shale. The lateral disappearance of several units, for example of the Hangenberg Black Shale, Drewer, and Hangenberg Sandstones, reflects a steep palaeoslope. Due to increasing turbulence upslope, siliciclastics did not settle towards the top of the seamount. The reworking in the Upper Crisis Interval, within the Pr. kockeli Zone, either reflects a minor regression or was the result of local synsedimentary tectonics, which triggered slumping. Glide folds are well-developed higher in the Carboniferous (Kronberg et al. 1960). GSSP Prospects: Drewer would be a prime GSSP candidate if the DCB was defined by the Hangenberg Black Shale. For a higher boundary, its advantages are a thick and complex Upper Crisis Interval. Its main disadvantages are the lateral pinching out of beds, the inaccessibility of the Stockum Level Black Shale and its important limestone nodules in the vertical quarry wall, and the scarcity of conodonts in the critical units. Above the Wocklum Limestone, many conodont taxa tend to be represented in normal-sized samples by single specimens only (see tables in . Therefore, it is hardly possible to study morphological changes within protognathids above the Hangenberg Sandstone. Apart from the top Wocklum Limestone, ammonoids are also rare. Kattensiepen (Fig. 7) Location: Variably active quarry at the road from Rüthen to Suttrop, topographic sheet 4516 Rüthen, r 3 457 980/h 5 703 750, northern limb of Warstein Anticline ( Fig. 9; asymmetric Kattensiepen Anticline). Literature: Staschen (1968), Struckmeier (1974), Stoppel 1977, Heuser et al. (1977, Clausen and Leuteritz (1979b), Clausen et al. 1984b), Luppold et al. (1994), , Koch et al. (2003: type locality of the arthropod Suttropcaris bottkei), Slotta et al. (2011). Summary of succession: The Famennian to basal Carboniferous consist of ca. 60 m, very strongly cyclic, well-bedded nodular limestone. Only the Annulata Event beds are rich in macrofauna, especially in ammonoids. Based on logging by H. Uffenorde, the DCB succession includes: Belecke Member of Eichenberg Formation: 2.5 m darkgrey shale with phosphatic nodules, especially at the base and above a short outcrop gap. Hangenberg Limestone: 70 cm middle-grey, nodular, argillaceous limestone with S. (Eo.) sulcata s.l./Pr. kuehni, Ps. triangulus inaequalis, and Ps. triangulus triangulus Zones in succession, post-Crisis Interval. ?Stockum Limestone: 40 cm silty limestone. Hangenberg Sandstone: 58 cm grey, calcareous, micaceous siltstone, Middle Crisis Interval II. Hangenberg Shale: 17 cm silty shale, Middle Crisis Interval I.
-short outcrop gap -Wocklum Limestone: Nodular and flaserlimestone. Discussion: Luppold et al. (1994) note that the strongly condensed DCB interval is represented by siliciclastics without macrofauna and relevant conodonts. Unconformities (gaps) are likely and the tectonic deformation is strong. This excludes the quarry from further DCB discussions.
Discussion: Attached to a discussion of his new genus Paragattendorfia, Schindewolf (1924) briefly introduced two new species of Gattendorfia. His Gattendorfia involuta was explained to differ from the type species of the genus (G. subinvoluta) by its more compressed and more narrowly umbilicated conch shape with higher whorls. This brief characterisation was sufficient at the time (before 1931) for a valid species definition (see Article 12 of the Code for Zoological Nomenclature). Statements in Weyer (1965aWeyer ( , 1972 and  that G. involuta is a nomen nudum are unjustified. Possibly influenced by Richter (1948; see discussion in Weyer 1972, p. 340), who wrongly advocated that an illustration was mandatory for a valid species definition, Schindewolf (1952, p. 298) assumed himself that his "third Gattendorfia species" from 1924 did not yet have a valid name. Therefore, he re-described it as G. tenuis, not mentioning at all his given name from 1924. He further stated that his previously available well-preserved specimens from Oberrödinghausen and Silesia, the involuta syntypes, had been lost. Therefore, he selected a specimen collected by Schwan from Gleitsch, Thuringia, as the holotype of G. tenuis (Tübingen collection Ce 1012/37). The neotype of G. involuta selected here is a well-preserved specimen from one of the two type localities (Oberrhödinghausen). Vöhringer (1960) and  illustrated several cross-sections of G. involuta topotypes (as G. tenuis), which enable a morphometric characterisation. Gattendorfia tenuis is a subjective junior synonym since it is based on a form from Thuringia, not from the Rhenish Massif. Its previously assumed identity with G. involuta has to be tested by investigations of ontogenetic morphometry. Two early whorl cross-sections of G. tenuis, illustrated by Weyer (1977) from the Schleiz region of Thuringia, suggest a gradual change from broadly depressed to higher whorls during the fifth whorl, possibly earlier than in typical G. involuta (during the 7th whorl in the involuta topotype Ce 1130/176, illustrated by Vöhringer 1960 and. This may reflect intraspecific variability but more morphometric data are required to solve the question. The new specimen illustrated in Fig. 15 is the so far oldest representative of the species from the basal G. subinvoluta Subzone (base of Bed 6a at Oberrödinghausen, G. subinvoluta Subzone, LC I-A 2 ). It suggests that G. involuta belonged to the root stock of the Gattendorfiinae.
A currently incomplete knowledge of morphology and distribution patterns is not only true for the pair G. involuta-G. tenuis. It also applies to G. subinvoluta, which holotype and a topotype from Gattendorf, Franconia (see Frech 1902;Schindewolf 1923a;, lack the constrictions that are well developed in Rhenish representatives figured by Vöhringer (1960), , Korn and Weyer (2003), and here (Fig. 3l, m). However, Vöhringer (1960, p. 151) noted that constrictions are absent in some specimens from Oberrhödinghausen and R. Richter collected in 1925 an adult from there (housed in the Senckenberg collection, Frankfurt a.M.) that has no constriction in the last ca. 250°o f the last whorl at 72 mm diameter. This suggests ontogenetic changes, with increasing spacing of constrictions towards maturity. Thuringian G. subinvoluta illustrated by Bartzsch and Weyer (1982) also lack prominent constrictions, which supports a possible difference between Rhenohercynian and Saxothuringian populations. These were biogeographically separated in the basal Carboniferous by a narrow oceanic system.

The Rhenish
Massif is one of the most important classical regions for DCB research, which successions set the standards for conodont, ammonoid, and event stratigraphy. A large part of the data was published only in regional journals or in German. Therefore, the initiative of this special issue to summarise internationally the DCB knowledge is a good opportunity to make this information better available. 2. The uppermost Famennian to basal middle Tournaisian interval can be subdivided into ten conodont zones, which are mostly defined by the FADs of species. As an exception, one zone, the costatus-kockeli Interregnum, is defined by the global extinction (LADs) of important taxa, such as Bi. costatus and Bi. ultimus (all subspecies), Ps. marburgensis trigonicus, and the last palmatolepids (all subspecies of Pa. gracilis, assuming a reworked nature of rare post-Hangenberg Extinction records). 3. The poor definition and ongoing revision of the oldest siphonodellids require to add currently the phrase s.l. (sensu lato) to identifications as S. (Eo.) praesulcata and S. (Eo.) sulcata. The type of S. (Eo.) praesulcata originated from a level above the main Hangenberg Extinction. It is likely that the most common and stratigraphically most useful pre-extinction siphonodellid will receive in future a different species name, with implications for zonal nomenclature. It is also possible that the use of the name S. (Eo.) sulcata eventually will be restricted to forms close to its lost holotype.
provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.