2.1. OVERVIEW OF GEOLOGY

Within Austria, weakly metamorphosed Paleozoic units are irregularly distributed throughout the Alpine orogen (Figure 1). Their original, pre-Alpine geographical location is uncertain, but it is generally accepted that the units were part of the northern Gondwana margin (e.g., ^{Neubauer et al., 2022}). According to tectonic concepts of the Eastern Alps (e.g., ^{Schmid et al., 2004}) - except for the Paleozoic remnants south of the Periadriatic fault -they are part of the uppermost tectonic nappes of the Eastern Alps, the Upper Austroalpine Nappe System. The Upper Austroalpine Nappe System consists of highly varied crystalline units with a dominant pre-Alpine metamorphic overprint, and is tectonically overlain by the Greywacke Zone, which itself is largely unconformably overlain by thick Mesozoic sediments of the Calcareous Alps. The Graz Paleozoic and some isolated outcrops in southern Burgenland which both lack Permo-Mesozoic cover sequences, as well as the Gurktal nappe, are part of the Drauzug-Gurktal Nappe System, which is itself part of the Upper Austrian Basement Nappes.

The structure of the Upper Austroalpine Nappe System was triggered by the closure of the Meliata Ocean (a marginal basin of the Tethys) during the Early Cretaceous. Therefore, the Upper Austroalpine Nappe System is sealed by Upper Cretaceous to Eocene sediments. Later, the Upper Austroalpine Nappe System was dismembered by lateral extrusion, i.e., eastward displacement of the eastern parts of the Eastern Alps (^{Ratschbacher et al., 1991}; ^{Neubauer et al., 2000}) and thus brought into its current position.

The internal architecture of the Graz Paleozoic (Silurian to Carboniferous) is composed of a high-grade metamorphosed lower nappe system and a low-grade metamorphosed upper nappe system (^{Gasser et al., 2010}; Figure 2). Each of them is differentiated into individual nappes. In short, the volcanosedimentary sequence contains basal pre-Devonian volcaniclastics, and a pronounced platform- to basinal-facies development during the Devonian. In the structurally higher units, especially in the Rannach Nappe, the depositional environment deepens steadily from the Upper Devonian to the Mississippian, and in the Pennsylvanian the depositional environment becomes abruptly shallow marine (Figure 3).

Resting on metamorphic basement, the nappes of the Graz Paleozoic are unconformably overlain by an Upper Cretaceous sequence, the "Kainach Gosau" (^{Ebner and Rantitsch, 2000}) in the west and by Neogene sedimentary rocks of the Styrian Molasse Basin in the south (^{Gross et al., 2007}).

In contrast to the Lower Nappe System which experienced stronger deformation as well as higher metamorphosis - up to lower amphibolite facies - sequences of the upper nappe system, among them the sequence of the Rannach Nappe, are quite fossiliferous.

2.2. RANNACH NAPPE OF THE GRAZ PALEOZOIC

The sequence of the Rannach Nappe (Figures 3, 4 and 5) starts with predominantly alkaline metavolcanites (tuffs, lavas) which pass into slightly calcareous sediments and contain fossils from the late Silurian to early Devonian (Ludfordian - Lochkovian). This is followed by various Lower Devonian (Lochkovian) to Upper Devonian (Frasnian) shallow-water platform carbonates. Formations of this depositional episode are comprised to the Rannach Group (^{Flügel, 2000}). Approximately during the Frasnian to Famennian transition, the depositional environment changed to an open, pelagic marine area. This persisted until the late Mississippian (Serpukhovian). The formations of the pelagic facies are combined into the Forstkogel Group (^{Flügel, 2000}). The end of the Rannach Nappe stratigraphic series is formed by lower Pennsylvanian (Bashkirian) shallow-marine birdseye limestones and phytoclastrich shales of the Dult Group (^{Flügel, 2000}). Detailed descriptions of formations of the Rannach, Forstkogel and Dult Groups, their lithologies, stratigraphic extents and the extensive literature can be found in the explanations for the "Austrian Stratigraphic Chart 2004" (^{Hubmann et al., 2013}).

The pelagic facies of the Forstkogel Group, which consists of micritic deep-marine cephalo-pod limestones of the Steinberg and Sanzenkogel Formations (Figures 4 and 5), reaches a maximum thickness of 100 m. In the Rannach Nappe east of the Mur (= Eastern Graz Paleozoic) it reduces down to about 30 m. The reasons for the reduced thickness are intraformational stratigraphic gaps in the area of the DCB, which were created by karstification (^{Ebner, 1978}, ^1980a, ^1980b).

Elementary geochemistry used for interpretions of bathymetry shows clear geographical differences: While in the Graz Paleozoic west of the Mur (= western Graz Paleozoic) the depositional depth in the Mississippian period was around 200 m, in the eastern Graz Paleozoic the water depth was less than 50 m (^{Nössing, 1974b}; ^{Ebner and Prochaska, 1989}; ^{Buchroithner et al., 1979}). Continuous sequences covering the DCB interval only occur in the western Graz Paleozoic. In the eastern Graz Paleozoic, however, erosion gaps occur in this time interval, with stratigraphic gaps increasing in temporal extent towards the east (Figure 3).

3.1. OVERVIEW OF LOCALITIES

Faunal and lithologic markers from different lithologic successions from the Graz Paleozoic, such as phosphorite nodules, lydites, fissure fillings, conodont index fossils and mixed conodont faunas, can be easily correlated (Figure 4) and resulted in the reconstruction of the depositional environments (see Section 2.2). The stratigraphic gap between the Steinberg Fm. and the Lower Sanzenkogel Fm. increases from west to east (Figure 3), and complete Famennian and Tournaisian sediments, including the DCB, are recorded at Forstkogel in the Steinberg area (^{Surenian, 1977}, ¹⁹⁷⁸; ^{Nössing, 1974b}, ^{Buchroithner et al., 1979}; ^{Ebner, 1980a}), and at Eichkogel in the Rein area (Figures 4 and 5; ^{Nössing, 1974a}; ^{Nössing et al., 1977}). Successions at Eichkogel are situated in a forest area and currently covered by forest soil. At Forstkogel, ^{Buchroithner et al. (1979)} described several quarries where DCB successions crop out, among them the famous Trolp quarry (Figure 6).

Complete successions of the Steinberg Fm. in the western Graz Paleozoic, reaching a thickness of 70 m, are reported at Forstkogel, Eichkogel and Weihermühle in the Gratberg-Au area (Figures 4 and 5; ^{Ebner, 1980a}, ^1980b). Complete successions of the Sanzenkogel Fm., reaching a thickness of up to 35 m from the Siphonodella sulcata to the Gnathodus bilineatus bollandensis conodont Zones, are reported at Forstkogel, while a hiatus is already reported from Weihermühle from the lower Tournaisian (Figure 4). The Tournaisian successions at Forstkogel were selected as type locality for the Lower Sanzenkogel Fm., established by ^{Nössing (1974b}, ¹⁹⁷⁵⁾. Sedimentologic investigations of the Lower Carboniferous (Mississippian) at Forst-kogel were summarized by ^{Ebner and Prochaska (1989)}.

In the western Graz Paleozoic, DCB successions at Trolp (see Section 3.2) and at Kanzelkogel (north of Graz, ^{Kodsi, 1967}) consist characteristically of condensed thin-bedded micritic limestones. At Kanzelkogel, the Upper Devonian and Lower Carboniferous (Mississippian) are separated by a several-decimetre-thick sequence of thin-bedded limestones with mixed Devonian and Carboniferous conodont faunas (^{Kodsi, 1967}, Figure 12). However, the Protognathodus or Siphonodella index conodont faunas were not recorded at Kanzelkogel probably due to a stratigraphic gap, or condensation and an incomplete sampling. Unfortunately, these successions do no longer exist due to mining activity.

3.2. TROLP SECTION (FORSTKOGEL, STEINBERG AREA)

The Trolp section is an abandoned quarry (ÖK Sheet 163 Voitsberg) located at Forstkogel about 8 km from the city Graz, beside the street from Steinberg to Rohrbach (Figures 5 and 6). Trolp is the type locality of the Late Devonian conodont Polygnathus styriacus first described by ^{Ziegler (1957)}. The Famennian-Tournaisian pelagic cephalopod limestones at Trolp have been intensively investigated due to rich conodont faunas, and sampling by Sandberg and Ebner in 1980 was followed by conodont studies of, for example, ^{Ebner (1980a)}, ^{Nössing (1975)}, ^{Kaiser 2005}, ^{Kaiser et al. (2008}, ²⁰²⁰⁾. It was formally proposed as a candidate for the GSSP at the DCB, but the scarcity of macro-fossils did not fulfill the criteria required (^{Ebner, 1980a}; ^{Ziegler and Sandberg, 1984a}; ^{Ji et al., 1989}).

The tectonically inverted (upside-down) successions at Trolp are dated as Late Famennian expansa-praesulcata Zone to Late Tournaisian Gnathodus typicus Zone (^{Ebner, 1980a}; ^{Kaiser, 2005}). A detailed, bed-by-bed or cm-by-cm sampling of DCB beds is required at Trolp due to the condensation of successions (Figures 7 and 8) which may produce a bias of the conodont record; renumbering of DCB beds was necessary due to weath ering and missing old numbers, as well as due to the high-resolution sampling of the successions. As shown in Figure 9, sample and bed numbers of ^{Kaiser (2005)} are correlated with bed numbers established by C. Sandberg and F. Ebner in 1980, and ^{Ebner (1980a)}. The praesulcata and kockeli Zones, and the sulcata/kuehni Zone (DCB level) are readily recognized by their zonal index fossils (^{Ebner, 1980a}, ^{Kaiser, 2005}; ^{Kaiser et al., 2009}, ²⁰²⁰; Figures 9 and 10a). The Middle praesulcata Zone of the standard conodont zonation (^{Ziegler and Sandberg, 1984b}), recognized at Trolp by Sandberg and Ebner in 1980, ^{Ebner (1980a)}, ^{Bojar et al. (2013)}, ^{Kaiser (2005)}, and by ^{Kaiser et al. (2008)} was not considered recently due to the asynchronous last occurrence of Palmatolepis gracilis gonioclymeniae defining the base of the Middle praesulcata Zone (see discussions in ^{Kaiser, 2005}; ^{Kaiser et al., 2009}; Corradini et al., 2016; ^{Spalletta et al., 2017}). The asynchronous last occurrence of this taxon resulted in different levels of the base of the Middle praesulcata Zone as previously recorded at Trolp by the above mentioned authors.

The eastern part of the quarry which is well accessible consists of well-exposed Lower Carboniferous beds, representing the lower member of the pelagic Sanzenkogel Formation ("Untere Sanzenkogel-Schichten"), and of bedded to massive upper Famennian beds, representing the Steinberg Formation. The succession consists of light-grey to brownish pelagic nodular cephalopod limestones and micritic Flaserlimestones of the Famennian Steinberg Fm. and Tournaisian Sanzenkogel Fm.; continuous limestone successions (Figures 7-10a and 10b) deposited during the Hangenberg Crisis interval are recorded at Trolp, which is globally characterized by black shales, sandstones or hiatuses. The stratigraphically lower part of the section consists of bedded to massive limestones of the praesulcata Zone, and a lithologic change is marked by a thin, ~1 cm thick shaly layer (basal Bed 10) at the base of thin-bedded limestones that disappears laterally (N 47°04'08'', E 15°19'09"). ^{Bojar et al. (2013}, Figure 3d) recorded a continuous limestone succession lacking this layer in a lateral, different position next to previous mentioned sampling points (N 47°04'29'', E 15°19'15''). This shaly or marly level was regarded by ^{Flügel and Ziegler (1957)} as a tectonic disturbance, which locally separated the Devonian and Carboniferous at Steinberg. Due to its stratigraphic position, ^{Kaiser (2005)} suggested an equivalent level of the Hangenberg Black Shale (HBS) since the lithologic change is accompanied by the major conodont biofacies change (Figure 11) indicating the main extinction phase of the Hangenberg Crisis at the costatus-kockeli Interregnum (ckI, after ^{Kaiser et al., 2009} = mass extinction-based level within the Middle praesulcata Zone, correlative with the HBSE and HSSE). This interpretation is supported by the onset of conodont radiations and the entry of Protognathodus meischneri, Protognathodus collinsoni, Protognathodus semikockeli and Protognathodus kockeli immediately above in thin-bedded limestone successions conformably overlying the shaly layer (^{Kaiser et al., 2020}; Figures 9 and 10a). The occurrence of stratigraphically leaked younger Carboniferous faunas - mainly siphonodellids and Gnathodus - in Devonian samples were recognized about at the level of this shaly layer, or in younger levels, respectively.

The base of the Carboniferous has been fixed at Trolp in Bed 16 by the first occurrence of index conodonts at the same stratigraphic levels (Figures 8, 9 and 12): Polygnathus purus subplanus, an index conodont for the base of the Tournaisian (e.g.,^{Kaiser et al., 2009}; ^{Spalletta et al., 2017}), as well as Siphonodella sulcata (M5) and Protognathodus kuehni, which were used to define the joint sulcata/kuehni Zone and the DCB (^{Kaiser et al., 2020}; ^{Becker et al., 2021}). Conodont index fossils previously recorded by ^{Ebner (1980a}) and ^{Nössing (1975)} at Trolp, and their correlation with index fossils recorded by ^{Kaiser (2005)} and ^{Kaiser et al. (2020)} are shown in Figure 9; the biostratigraphic corrrelation of the Hangenberg Event and the DCB from the Carnic Alps and Rhenish Massif is shown in Figure 8. At stratigraphically younger levels, Siphonodella bransoni (Siphonodella duplicata M1), as well as Pseudopolygnathus multistriatus were recorded (^{Ebner 1980a}). The condensed Tournaisian sequence is intercalated by the 20 cm thick Trolp Phosphorite Bed (Figures 4 and 14), marking the beginning of the Upper Sanzenkogel Fm. (base of the upper Tournaisian) after ^{Nössing (1975)}, or the base of the middle Tournaisian as suggested by ^{Kaiser (2005)}, respectively. It is correlative with the end of the erosional gap within parts of the Rannach Nappe, and interpreted as deepening and upwelling at the shelf margin at the base of the Scaliognathus anchoralis Zone (base of upper Tournaisian; ^{Ebner, 1998}; ^{Ebner et al., 2000}).

Conodont biofacies at Trolp (Figure 11) indicates a dominance of Bispathodus, Branmehla and Palmatolepis in pre-Hangenberg Crisis (pre-extinction) levels in the praesulcata Zone (^{Kaiser 2005}). The conodont mass extinction in the ckI at Trolp (base of Bed 10), and major biofacies change to an impoverished Polygnathus-Protognathodus biofacies with stunted faunas, is connected to the extinction of Famennian taxa which predominated in the Famennian, such as Branmehla suprema, Bispathodus costatus, Palmatolepis gracilis expansa, and Pseudopolygnathus marburgensis trigonicus. The beginning of radiations in the basal kockeli Zone, recorded in the basal part of the thin-bedded limestones (Bed 10/11), is succeeded by faunal recovery and the 2nd radiation phase (Figure 10a) in the basal Lower Carboniferous sulcata/kuehni Zone. It is connected to the return of normal-sized conodont faunas, and an increasing abundance of pseudopolygnathids and siphonodellids, but the Polygnathus-Protognathodus biofacies remains in the Lower Carboniferous.

Faunal evolution during the Hangenberg Crisis is precisely documented at Trolp (^{Kaiser et al., 2020}). A faunal assemblage of the Lower Protognathodus fauna, with Protognathodus meischneri and Protognathodus collinsoni, enters at a level with the initial litho- and biofacies change (Bed 10, Figures 9, 10a and 11), and the absence of Bispathodus costatus and Protognathodus kockeli at the same level in Bed 10 is interpreted as the event-based ckI. The succeeding entry of Protognathodus semikockeli recently established by ^{Hartenfels et al. (2022)} occurs immediately above, and the entry of Protognathodus kockeli s.str. slightly above marks the start of the kockeli Zone (Figures 9 and 10a). The Protognathodus fauna at the DCB at Trolp even occurs in association with Siphonodella praesulcata and Siphonodella sulcata which was also recorded by ^{Sandberg et al. (1983)}.

At Trolp, a phyletic lineage of the early protognathodids (Protognathodus meischneri, Protognathodus collinsoni, Protognathodus semikockeli, Protognathodus kockeli s.str., advanced Protognathodus kockeli, Protognathodus kuehni (see figure 6 in ^{Kaiser et al. 2020}) as proposed by ^{Ziegler (1973)}, ^{Ziegler and Leuteritz (1970)} and ^{Hartenfels et al. (2022)} is recognized on the basis of the shape and the ornamentation of the cup of typical morphotypes (Figures 12 and 13). Many specimens, however, can be designated as atypical morphotypes (Figure 13; see ^{Kaiser et al., 2019}unpublished) recorded by ^{Nössing (1975)}, ^{Ebner (1980a)} and ^{Kaiser et al. (2020)}.

Stable isotopes (δ¹³C) at Trolp (^{Kaiser et al. 2008}; Figure 10b), used for the reconstruction of changes in the marine dissolved inorganic carbon reservoir, revealed a positive excursion in δ¹³C_carb in the ckI and kockeli Zone to values of up to 3‰ (V-PDB). Two minor positive 8¹³C_carb peaks were recorded in the praesulcata Zone, with values reaching 2.5‰ and 2.7‰, and one positive peak was measured in the Upper expansa Zone (Bispathodus ultimus Zone after ^{Becker et al., 2021}). The onset of continuously decreasing carbon isotope values is at the base of the sulcata/kuehni Zone (DCB). A δ¹³C_org excursion was measured from the sedimentary organic matter at Trolp, and oxygen isotope values of conodont apatite at Trolp are remarkable high indicating low seawater temperatures (unpublished data). Element geochemistry conducted by ^{Bojar et al. (2013)} at Trolp indicates a decrease of detrital influence in the lower Tournaisian and was interpreted as transgression.

4.1. LITHO,- BIO, -AND CHEMOSTRATIGRAPHY AT TROLP AND CORRELATIONS OUTSIDE THE REGION

Characteristic markers and abundant conodonts at Trolp allow a high time-resolution study of major environmental changes around the DCB. These can be well correlated with coeval markers from continuous limestone successions as well as siliciclastic-dominated successions in North-America, Asia, Europe, North-Africa from pelagic, hemipelagic and neritic realms as discussed as follows, indicating environmental changes on a global scale.

4.1.1. LITHOSTRATIGRAPHYAND LITHOFACIES

At Trolp, a characteristic lithofacies change (Figure 7), from bedded or massive limestones to thin-bedded limestones (mud- to wackestones, ^{Kaiser 2005}), is correlative with the global change to siliciclastics (HBS, HSS) and the anoxic episode of the Hangenberg Crisis in the extinction-based ckI and kockeli Zone (Figure 8-10). In continuous, pelagic limestone successions in the Carnic Alps (Grüne Schneid, Figure 8), and in hemipelagic to pelagic, more proximal successions in the Montagne Noire (La Serre É, e.g.,^{Flajs and Feist, 1988}; ^{Feist et al., 2021}) and in the Moravian Karst (^{Kumpan et al., 2021}), a lithofacies change from bedded to thin-bedded, condensed limestones can be correlated with equivalent changes at Trolp. This lithofacies change is globally recorded, and correlative for example with black shales and/or sandstones in pelagic but more deeper settings in the Carnic Alps (Kronhofgraben), Pyrenees (Milles), as well as in hemipelagic settings in the Rhenish Massif (Figure 8), for example at Drewer, or in neritic successions in the Meseta and Anti-Atlas, Morocco (Ain Jemaa, El Atrous, M'Karig, Lambidia), and in the Alborz Mts., Iran (Mighan, Chelcheli), as reported by several authors (e.g.,^{Cygan and Perret, 2002}; ^{Kaiser, 2005}; ^{Kaiser et al., 2006}, ²⁰⁰⁷, ²⁰¹¹, ²⁰¹³, ²⁰¹⁸; ^{Becker et al.,, 2016b}; ^{Bahrami et al., 2021}; ^{Parvizi et al., 2021}; ^{Konigshof et al., 2021}).

The lithofacies change was caused by a global carbonate crisis at the end of the Devonian (^{Li et al., 2022}) due to ceased carbonate sedimentation, enhanced C_org burial and marine anoxia during the Hangenberg Black Shale Event (HBSE), and the succeeding "Hangenberg glaciation" during the Hangenberg Sandstone Event (HSSE; e.g.,^{Isaacson et al., 2008}; ^{Kaiser et al., 2006}; ^{Lakin et al., 2016}). The main regressive phase which led to the global deposition of Hangenberg Sandstone equivalents cannot be recognized by a clear sedimentological marker at Trolp, while the onset of the anoxic episode may be represented by the thin shaly layer (base of Bed 10), which is, however, only locally developed.

Tournaisian cherty sediments at Trolp (Trolp Phosphorite Bed, Figures 4 and 14), can be correlated with the basal middle Tournaisian as suggested by ^{Kaiser (2005)}, and with the faunal turnover during the Lower Alum Shale Event which is globally characterized by a lithologic change from limestones to cherty rocks, the equivalents of the Rhenish Lower Alum Shale of the Middle crenulata Zone (^{Becker, 1993a}, ^1993b; ^{Kaiser et al., 2018}). It is coeval with transgressive shales, partly associated with phosphorite nodules, for example at Kronhofgraben, Carnic Alps (^{Schönlaub, 1969}; ^{Kaiser, 2005}), Puech de la Suque and La Serre É, Montagne Noire (^{Lethiers and Feist, 1990}; ^{Kaiser, 2005}; ^{Feist et al., 2021}), and at El Atrous, Jebel Ouaoufilal, M'Fis, M'Karig, Anti-Atlas, Morocco (^{Kaiser et al., 2011}, ²⁰¹³, ²⁰¹⁸). The base of the upper Tournaisian anchoralis Zone is globally characterized by an end of an erosional gap, deepening and upwelling. It is correlative with cherty sediments all over the Graz Paleozoic and with the end of an erosional gap within the Rannach Nappe, interpreted as deepening and upwelling event by ^{Ebner (1998)} and ^{Ebner et al. (2000)}.

4.1.2. BIOSTRATIGRAPHY AND CONODONT BIOFACIES

A continuous record of biofacies change at the DCB can be well recognized at Trolp, as well as in continuous limestones in the Carnic Alps, Grüne Schneid (Figure 11), and there is no hiatus in each phases of the Hangenberg Crisis. The dominance of the Protognathus fauna and the Polygnathus purus group in the kockeli and sulcata/kuehni Zones at Trolp, Grüne Schneid and elsewhere characterizes the 1st (FO of Protognathodus kockeli, = base of kockeli Zone) and 2nd (FO of Protognathodus kuehni, FO of Siphonodella sulcata M5, = basal Tournaisian transgression) faunal recovery phases after the Hangenberg extinction (Figure 14). The 3rd radiation phase is in the middle sulcata/kuehni Zone (FO of Siphonodella sulcata M4, Gattendorfia level = Tournaisian transgression).

At Trolp, the Hangenberg extinction-based ckI is recognized by the absence of Bispathodus costatus and Protognathodus kockeli at same level, and the extinction of almost all upper Famennian conodont taxa among the pseudopolygnathids, polygnathids, bispathodids, branmehlids and palmatolepids (see ^{Kaiser et al., 2009}, ²⁰²⁰). It is recognized by a major biofacies change from a Famennian Bispathodus-Branmehla-Palmatolepis biofacies to a characteristic, impoverished latest Famennian and earliest Tournaisian Polygnathus-Protognathodus biofacies. The ckI and the major conodont biofacies change are recorded at the base of the thin-bedded limestones at Trolp, as well as in continuous pelagic and hemipelagic limestone successions at the base of thin-bedded limestones in the Carnic Alps at Grüne Schneid (^{Kaiser, 2007}; Figure 11), in the Moravian Karst at Krtny (^{Kumpan et al., 2021}), and in the Montagne Noire at La Serre É (^{Flajs and Feist, 1988}; ^{Feist et al., 2021}, Kaiser unpublished data) at correlative levels, and is caused by major environmental changes during the HBSE and HSSE.

Conodont faunal distributions and lithofacies at Trolp display distal, pelagic depositional environments, with sediment deposition in ca. 200 m water depth (e.g.,^{Nössing, 1975}). For comparison, litho- and biofacies, and the depositional environments in the Carnic Alps at Grüne Schneid (^{Schönlaub et al., 1988}, ¹⁹⁹¹) are similar to that of Trolp, and a rich Protognathodus fauna occurs in both regions (^{Kaiser et al., 2020}). Differences only related to single occurrences of Protognathodus specimens in pre-Hangenberg extinction levels (praesulcata Zone) at Grüne Schneid, with minor differences in biofacies, only. Accordingly, a slightly higher abundance of Bispathodus than of Palmatolepis occurs in the Famennian in pre-crisis levels at Trolp (Figure 11) which is an indicator for more shallow depositional environments according to previous biofacies models (see ^{Kaiser et al. 2017}). For example, Bispathodus abundantly occurs in neritic settings in Iran (Alborz Mts., e.g., Mighan, Chelcheli), Morocco (Tafilalt, e.g., El Atrous), and in hemipelagic settings in the Rhenish Massif (e.g., Hasselbachtal), Montagne Noire (e.g., La Serre, Puech de la Suque), Moravian Karst (e.g., Lesni Lom), as reported by several authors (e.g.,^{Kaiser, 2005}; ^{Hartenfels and Becker, 2016}; ^{Feist et al., 2021}; ^{Kaiser et al., 2011}, ²⁰¹⁸; ^{Bahrami et al., 2011}, ²⁰²¹; ^{Konigshof et al., 2021}; ^{Kumpan et al., 2021}; ^{Parvizi et al., 2021}).

In contrast, siphonodellids, as indicator for more deeper environments (see ^{Kaiser et al., 2017}) are slightly more abundant at Trolp in pre-event levels when compared to Grüne Schneid, and Siphonodella is not recorded in the ckI and kockeli Zone at Grüne Schneid and elsewhere in the Carnic Alps. Protognathodids, previously suggested as an indicator for more shallow environments, occur at Grüne Schneid in pre-event levels, while this faunal group is not recorded in pre-event levels at Trolp and elsewhere in the Graz Paleozoic. In summary, faunal distribution of Protognathodus and Siphonodella indicates more shallow depositional environments at Grüne Schneid than at Trolp in pre-event levels, which is, however, not in accordance with the Bispathodus biofacies at Trolp as mentioned above.

In this respect, previous biofacies concepts for example of ^{Sandberg and Ziegler (1979)}, ^{Dreesen et al. (1986)} and ^{Ziegler and Weddige (1999)}, have to be reconsidered (see ^{Kaiser et al., 2017}, ²⁰²⁰). Paleogeographic provincialism and migration pattern, competition and feeding (^{Mossoni et al., 2015}), but also decreasing ecological niches around the DCB due to global shelf anoxia (^{Algeo et al., 1995}; ^{Li et al., 2022}) and major eustatic sea-level fall with longterm and short-termed regressions (e.g.,^{Johnson et al., 1985}; ^{Kaiser et al., 2011}; ^{Babek et al., 2016}), as well as climate cooling (Issacson et al., 2008; ^{Brezinski et al., 2009}; ^{Lakin et al., 2016}), caused more likely characteristic faunal assemblages and biofacies changes. The occurrence of either Protognathodus or Siphonodella around the DCB is thus more likely related to major environmental changes during anoxic, transgressive and and regressive phases of the Hangenberg Event, and is regarded as biotic opportunism by several authors (e.g., ^{Kaiser, 2005}, ^{Kaiser et al., 2008}, ²⁰⁰⁹; ^{Corradini et al., 2011}; ^{Spalletta et al., 2017}). Thus, disregarding taxonomic uncertainties as discussed below, the occurrence of Protognathodus in pre-Hangenberg Event levels in the Carnic Alps and other regions in Europe, Asia and North-Africa could thus be linked with an opportunistic lifestyle during environmental changes probably related to the Drewer Event (early Hangenberg phase) in the praesulcata Zone (see Section 4.1.3).

4.1.3. STABLE ISOTOPES AND CHEMOSTRATIGRAPHY

Chemostratigraphy as a formal stratigraphic method is a useful tool for global correlations of chronological boundaries (see ^{Sial et al., 2019}), correlations of event-related successions (e.g.,^{Algeo and Shen, 2024}), and Hangenberg eventrelated successions at the DCB (^{Kaiser et al., 2016}). In this respect, positive carbon isotope excursions recorded at Trolp from different levels in the Upper expansa (ultimus Zone) andpraesulcata Zones, as well as in the ckI and kockeli Zone (Figure 10b) can be well correlated outside the region (Figure 10c), and thus indicate enhanced marine C_org burial and global perturbations of the cabon cycle. Accordingly, a minor positive peak in δ¹³C_carb in the Upper expansa Zone at Trolp, with increasing values of more than 2‰, is widely known from the Middle (costatus Zone) and Upper expansa Zone as distinctive positive carbon isotope excursions of up to 3‰ in the Carnic Alps (^{Kaiser et al., 2008}), Franconia, southern Germany (unpublished data,^{Kaiser et al., 2022}), and the Moravian Karst (unpublished data,^{Kumpan et al., 2018}), and were correlated previously with the small-scaled "Epinette" (Middle expansa) and "Etreoungt" (Upper expansa) bioevents (^{Kaiser et al., 2008}). Increasing δ¹³C_carb values and positive carbon isotope excursions in shallow-water successions were also recorded from the Middle and Upper expansa Zone in South China (costatus and ultimus Zone, ^{Zhang et al., 2019}), in northern Iran (^{Parvizi et al., 2021}), and in Belgium (^{Kumpan et al. , 2014b}). In deeper pelagic settings (Kronhofgraben, Carnic Alps), a positive δ¹³C_carb spike of 2‰ is recorded as well in the middle part of the Middle expansa Zone (^{Kaiser et al., 2008}). Worldwide, transgressive shale and black shale deposits (^{Algeo et al., 1995}; ^{Li et al., 2022}; ^{Sahoo et al., 2023}), and enhanced C_org burial episodes, as also indicated by an increasing δ¹³C_carb baseline on a global scale (^{Zhang et al., 2019}), are recorded in the expansa Zone probably connected with the Dasberg Crisis in the Lower and basal Middle expansa Zones (see ^{Hartenfels, 2011}), and with the Epinette and Etreoungt Events in the middle Middle and Upper expansa Zone (e.g.,^{Kaiser et al., 2008}; see also ^{Di Pasquo et al., 2019}). Two minor positive δ¹³C_carb peaks in the praesulcata Zone at Trolp, reaching values of more than 2.5‰, can be correlated with coeval positive peaks and/or increasing carbon isotope values of up to 3‰ recorded from the Carnic Alps at Grüne Schneid, Rio Boreado, Casera Malpasso, Groβer Pal (Figure 10c A; ^{Kaiser et al., 2006}, ²⁰⁰⁸), and in Franconia (^{Kaiser et al., 2022}, unpublished data), as well as by positive C_org peaks in the Rhenish Massif (Figure 10c B; ^{Kaiser et al. 2006}). These geochemical anomalies were interpreted as anoxic phases of the Drewer Event during the early phase of the Hangenberg Crisis in the praesulcata Zone (^{Kaiser et al., 2020}). The Drewer Event was previously correlated with the regressive Drewer Sandstone and equivalents elsewhere (^{Streel, 1999}; see also ^{Kaiser et al., 2011}, ²⁰¹⁶), and was previously regarded as a short-term glaciation episode during a humid event at the LL-LE miospore transition (e.g.,^{Streel et al., 2000}). The Drewer level in the Rhenish Massif is characterized by at least two distinctive shale horizons (^{Becker et al., 2021}). Wet climate and enhanced detritus influx ( = regressive Drewer Sandstone Event) probably caused enhanced marine bioproductivity and suceeding global shelf anoxia already in the praesulcata Zone, well before the HBSE, and may expressed by enhanced carbon burial as indicated by the positive δ¹³C_carb peaks.

A significant increase of δ¹³C_carb values by almost 2‰ at Trolp is at the base of the thin-bedded limestones associated with the conodont mass extinctions and the major conodont biofacies change in Bed 10 as discussed in Section 4.1.2. The positive δ¹³C Hangenberg-excursion in the ckI and kockeli Zone of values up to 3‰ was globally recorded (Figure 10c, e.g.,^{Kaiser et al., 2006}, ²⁰¹⁶; ^{Cramer et al., 2008}; ^{Kumpan et al., 2014a}; ^{Matyja et al., 2021}; ^{Qie et al., 2021}; ^{Becker et al., 2021}). The onset of the positive excursion indicates global changes of the carbon cycle during the Hangenberg Crisis which was caused by enhanced C_org burial during the deposition of the HBS and equivalents (e.g.,^{Kaiser et al., 2006}). Enhanced bioproductivity, high rates of continental weathering and terrestrial nutrient influx due to erosion of organic matter during an upper Famennian sea-level lowstand (^{Babek et al., 2016}) contributed to enhanced organic carbon burial. This finally resulted in a succeeding major, glacio-eustatic sea-level fall during wet and cold climate, time-equivalent to the deposition of the regressive Hangenberg Sandstone and global equivalents, and a glaciation pulse (Hangenberg glaciation) at the end of the Famennian (e.g.,^{Kaiser et al., 2006}; ^{Isaacson et al., 2008}; ^{Brezinski et al., 2009}; ^{Lakin et al., 2016}).

A distinctive decrease of δ¹³C_carb values, starting in the lower Tournaisian sulcata/kuehni Zone at Trolp, reflects an overall long-term trend in the lower Tournaisian at same levels in numerous settings in European, Asian and North-American regions (Figure 10c, ^{Kaiser et al., 2016}, and references therein). This trend may be explained by a reduced sedimentation rate of marine organic matter, or by an enhanced delivery of light carbon into the oceans from continental weathering of organic-rich sediments during a short-term warming episode in the basal Tournaisian (^{Kaiser et al., 2006}; ^{Marshall et al., 2020}). Decreasing carbon isotope values at Trolp were reported as well by ^{Bojar et al. (2013)}. However, their data and interpretations are related to uncorrect positions of the kockeli Zone (= Upper praesulcata Zone) and the DCB (see ^{Bojar et al. 2013}, Figure 3a). The DCB is fixed by the authors by the late FO of Siphonodella sulcata previously recorded by ^{Ebner (1980a)} from his Bed 8 (see Figure 9). The authors ignore that the DCB was correctly fixed by ^{Ebner (1980a)} in his Bed 4 with the FO of Polygnathuspurus subplanus as an index fossil for the DCB (^{Kaiser et al., 2009}; ^{Spalletta et al., 2017}). Bed 4 of Ebner is correlative with Bed 16 of Kaisers studies, according to thickness of the limestones and the FO's of Polygnathus purus subplanus, Protognathodus kuehni and Siphonodella sulcata (M5) in Bed 16. Thus, main parts of the kockeli Zone and the DCB of ^{Bojar et al. (2013)} can be regarded as younger, Tournaisian successions, and these biostratigraphic misinterpretations resulted in misinterpretations of geochemical proxies at Trolp (see ^{Zhang et al., 2019}, Figure 3). Further, the transgressive episode as interpreted from element geochemistry, and regarded as DCB-level by ^{Bojar et al. (2013)}, is more likely the expression of the lower Tournaisian transgression (Gattendorfia level) in the middle sulcata/kuehni Zone (sulcata Event after ^{Kalvoda and Kukal, 1987}).

4.2. SCENARIOS FOR A POTENTIAL GSSP POSITION AT THE DCB

4.2.1. AN EVENT-BASED SCENARIO

The event-based, early phase of the Hangenberg Crisis, recognized at Trolp by bio,- litho- and chemostratigraphy (Section 4.1), is the initial main Hangenberg mass extinction Event among many marine organisms (e.g., ^{Kaiser et al., 2016}, Figure 3). It is correlative with the global deposition of the transgressive Rhenish HBS (= HBSE) and its equivalents elsewhere, and the main anoxic phase of the Hangenberg Crisis, indicated by the first significant positive peak in δ¹³C_carb, δ¹³C_org (e.g., ^{Kaiser et al., 2006}). The global change from cephalopod limestones to black shales, or to high-condensed, thin-bedded limestones, has a worldwide correlation to numerous pelagic, hemipelagic and neritic settings in Asia, America and Europe (e.g.,^{Kaiser et al., 2016}; ^{Becker et al., 2016a}, ²⁰²¹; ^{Aretz et al., 2021}, ^{Kumpan et al., 2021}; ^{Matyja et al., 2021}, ^{Konigshof et al., 2021}; ^{Komatsu et al., 2014}; ^{Qie et al., 2021}).

The HBSE (former middle part of the Middle praesulcata Zone) can be recognized biostratigraphically for example by miospores, and is dated as the LN miospore Zone (^{Sandberg et al., 1972}; ^{Streel, 2009}). The base of the LN miospore Zone provides correlation into the terrestrial realm and into macrofossil-free high-latitude settings. Furthermore, this level can be recognized by sequence stratigraphy and the drowning unconformity at the base of a maximum flooding interval (e.g., ^{Van Steenwinkel, 1993a}, ^1993b; ^{Kaiser et al., 2011}). After summaries in ^{Becker et al. (2016a)}, this episode of worldwide marine anoxia can be geochemically determined in various regions by organic geochemistry, inorganic carbon isotopes, magnetosusceptibility, element geochemistry, and by gamma ray spectroscopy (e.g.,^{Marynowski and Filipiak, 2007}; ^{Kaiser et al., 2006}; ^{Cramer et al., 2008}; ^{Kumpan et al., 2014a}, ^2014b, ²⁰¹⁵).

In siliciclastic-dominated successions, it is unclear which taxa range through the anoxic and regressive phase (HBSE, HSSE) due to fossil poor sediments, while in continuous pelagic limestone successions at Trolp and Grüne Schneid, the ckI (HBSE, HSSE) marks an important level. This level is characterized by main conodont extinctions, decreasing number of faunas, poor and stunted faunas, and major biofacies change. Alternative zonation concepts (Figure 12) do not consider the main Hangenberg extinction level instead, the level of the ck is the top part of the ultimus Zone (^{Spalletta et al., 2021}), or the base of the extended kockeli Zone (^{Corradini et al., 2017}).

The HBSE was previously proposed as a natural DCB (^{Walliser, 1984}). However, an event-based scenario is difficult to establish because 1. event-related successions are often accompanied by leakage, reworking, and impoverished faunas, and 2. hiatuses and C_org burial phases well before the HBSE, connected with the Dasberg, Etreoungt and Drewer Events as discussed in Section 4.1.3, resulted in the asynchronous onset of global black shale depositions and stratigraphic gaps (see figure 2a in ^{Bojar et al., 2013}).

4.2.2 THE BASE OF THE KOCKELI ZONE

The base of the kockeli Zone (= Upper praesulcata Zone), as well as the extended kockeli Zone (= ckI/kockeli Zone to base of the bransoni Zone, see Figure 12) are not applicable as a potential DCB position as discussed as follows. The start of the kockeli Zone is probably within the upper LN miospore Zone, but the correlation into the terrestrial realm is difficult (e.g., ^{Becker et al., 2016a}). The entry of Protognathodus kockeli in its former definition of ^{Bischoff (1957)} - related to diagnostic features as curvature (bending) and shape of the cup as well as platform ornamentation - is affected by the restricted distribution of Protognathodus kockeli in many regions and widespread facies-controlled, discontinuous cono-dont sampling across the base of the kockeli Zone due to carbonate-poor deposits and late entry of the index fossil with the first carbonates (see ^{Kaiser et al., 2020}). The kockeli Zone is commonly not preserved due to the absence of Protognathodus kockeli at this level, as previously reported and additionally discovered recently in the Alborz Mts., Iran (e.g., Mighan, Chelcheli; ^{Bahrami et al., 2021}; ^{Königshof et al., 2021}; ^{Parvizi et al., 2021}), Carnic Alps (e.g., Kronhofgraben, Plan di Zermula; ^{Kaiser et al. 2009}; ^{Spalletta et al. 2021}), Anti-Atlas (e.g., M'Karig, El Atrous; ^{Kaiser et al., 2011}), Moravian Karst (Krtny; ^{Kumpan et al., 2021}), or due to an impoverished Polygnathus-Protognathodus biofacies at the end of the Famennian (partly without Protognathodus kockeli), respectively, for example in the Rhenish Massif (e.g., Oese, Oberrõdinghausen; ^{Kaiser et al., 2017}; see ^{Becker et al., 2021}), Holy Cross Mountains (Kowala; ^{Matyja et al., 2021}), Illinois Basin, Indiana, USA (^{Over, 2021}). Stratigraphic gaps are caused by the reduced formation of carbonates due to a global carbonate crisis during this time span, resulting from continuing Famennian sea-level fall (^{Babek et al., 2016}) as the longterm regression continued until the Carboniferous, and from continuing major global cooling (e.g., ^{Isaacson et al., 2008}; ^{Brezinski et al., 2009}; ^{Kaiser et al., 2022}, unpublished data). The absence of the lower Protognathodus conodont assemblages (Protognathodus meischneri, Protognathodus collinsoni, Protognathodus kockeli) does not necessarily suggest stratigraphical gaps because erosional surfaces are missing (e.g., Rhenish Massif, Iran and elsewhere, see ^{Becker et al., 2021}; ^{Königshof et al., 2021}) but is caused by biofacies changes and the deteriorating environment after the early and middle Hangenberg phase.

The entry of Protognathodus kockeli could be affected by local fossil reworking or leakage, for example in the Carnic Alps (^{Corradini et al., 2017}; see discussions in ^{Kaiser et al., 2020}). Reworking and mixing of conodonts at this level is the result of the major eustatic fall associated with the HSSE and the Hangenberg glaciation in the middle Hangenberg phase, while leakage is the result of paleokarstic phenomena in the Graz Paleozoic and Carnic Alps (e.g.,^{Ebner, 1980b}; ^{Schönlaub et al., 1991}).

Also, the uncertainties of the FAD of the "one-row" (Protognathodus semikockeli after ^{Hartenfels et al., 2022}) and the "two-rows" morphotype of Protognathodus kockeli (Protognathodus kockeli s.str. after ^{Hartenfels et al., 2022}), and the start of the kockeli Zone either during the middle (regressive episode of the HSSE, based on stable isotopes and the FO of Protognathodus semikockeli and Protognathodus kockeli at Trolp and Grüne Schneid, ^{Kaiser et al., 2022} unpublished data), or late phase of the Hangenberg Crisis (after the facies break related to the HSSE (^{Hartenfels et al., 2022}) have to be considered. The base of the kockeli Zone (former Upper praesulcata Zone) was previously established by the "two-rows" Protognathodus kockeli (^{Bischoff, 1957}), while it was established by the "one-row" Protognathodus kockeli by ^{Corradini et al. (2011)}. Based on conodont studies at Borkewehr (Rhenish Massif; ^{Hartenfels et al., 2022}) from limestone successions overlying the regressive Hangenberg sandstone equivalent, the FO of Protognathodus kockeli s.str. (two-rows) within the evolutionary lineage of Protognathodus semikockeli (one-row) and Protognathodus kockeli s.str. in the postregressive Hangenberg phase is proposed as future GSSP for the DCB.

4.2.2.1. CONSIDERATIONS OF PROTOGNATHODUS

Taxonomic and biostratigraphic uncertainties within typical morphotypes of the early Protognathodus faunas are known for many years and were rementioned in ^{Corradini et al. (2011)}, ^{Kaiser et al. (2019} unpublished, ²⁰²⁰⁾ and ^{Hartenfels et al. (2022)}. Originally, ^{Bischoff (1957)}, ^{Ziegler and Leuteritz (1970)}, and ^{Ziegler (1969)} only considered randomly the high spectrum of Protognathodus morphotypes. Generally, the high spectrum of intermediate morphotypes of Protognathodus species attests to the completeness of the stratigraphic record, as at Trolp, Grüne Schneid (^{Kaiser et al., 2020}), Puech de la Suque (France) and Borkewehr (^{Hartenfels et al., 2022}). However, taxonomic uncertainties are related to intraspecific variability concerning juvenile-adult and early-advanced forms, as well as intermediates between all species of the early Protognathodus fauna (^{Kaiser et al., 2019}, ²⁰²⁰; ^{Hartenfels et al., 2022}). This was previously recognized in the early siphonodellids, too; the high spectrum of intermediate morphotypes of Siphonodella praesulcata and Siphonodella sulcata at La Serre É and elsewhere have been taxonomically reevaluated (^{Kaiser and Corradini, 2011}), and morphotype groups of both Siphonodella praesulcata and Siphonodella sulcata were established and widely used (Figure 12, see Section 4.2.3.1).

Diagnostic features of the Protognathodus fauna are related to the cup morphology (curvature and shape of the cup) and platform ornamentation (^{Bischoff, 1957}; ^{Ziegler and Leuteritz, 1970}), and a phyletic lineage of the early protognathodids (^{Ziegler, 1973}), based on the typical morphotypes and their transitional forms (intermediates), was suggested at Trolp and Grüne Schneid (^{Kaiser et al., 2020}), comparable to Puech de la Suque from Bed 10 overlying the HBS (^{Hartenfels et al., 2022}). Accordingly, a well-exposed succession of Protognathodus meischneri, Protognathodus collinsoni, Protognathodus semikockeli, Protognathodus kockeli and Protognathodus kuehni could be recorded in these two regions. After ^{Kaiser et al. (2020)}, the typical morphotype of Protognathodus kuehni at Trolp has at least on one side of the cup distinctive transverse ridges running radial from the platform edge to the carina, and the shape of the cup is wide and slightly asymmetric according to the original diagnosis of ^{Ziegler and Leuteritz (1970)}.

Three new Protognathodus species are defined based on the arrangement of the cup ornamentation which focused on the developing of row of nodes (^{Hartenfels et al., 2022}), without considering the shape and curvature of the cup, and the platform ornamentation "scattered nodes' which are discussed by the authors, only (see also ^{Corradini et al., 2011}, and ^{Kaiser et al., 2020}). Transitional forms, transitional forms between Protognathodus kockeli and Protognathodus kuehni, but also morphotypes previously determined as Protognathodus kuehni are included in Protognathodus kockeli s.str. (^{Hartenfels et al., 2022}), and thus enlarged the spectrum of morphotypes to be determined as Protognathodus kockeli.

Atypical morphotypes of "Protognathodus' species with a high morphological variability, related to the shape of cup and the ornamentation of the platform, are recorded in the Graz Paleozoic (Figure 13), Carnic Alps, and other regions (see ^{Kaiser et al., 2020}; ^{Corradini et al., 2011}, ^{Hartenfels et al., 2022}), and as stated in ^{Hartenfels et al. (2022)}, "... atypical characteristics [within the genus Protognathodus], ... result in a significant morphological complexity". At Trolp, within all species of Protognathodus, typical and also atypical morphotypes at Trolp are observed at the same stratigraphic levels in the uppermost Famennian and lower Tournaisian (^{Nössing, 1975}; ^{Kaiser et al., 2020}) which were partly included in the three new Protognathodus species (Protognathodus semikockeli, Protognathodus kockeli s.str., Protognathodus kockeli) recently established (^{Hartenfels et al., 2022}). For example, data at Trolp (Figure 13) indicate that beside specimens with a Protognathodus-like shape of cup, instead extremely narrow, atypical (instead of wide in comparison to holotypes) outlines of the shape of the cup in all species of the protognathodids are recorded from the kockeli to bransoni Zones (^{Kaiser et al., 2020}). Other morphologic characteristics are an extended posterior tip, a strong asymmetry due to an extended outer side of the cup, an enlarged outer margin which is commonly broken, or a strongly bent carina, in contrast to specimens with a straight carina.

Similar to the group of "siphonodelloids', which has morphological features of Siphonodella, Polygnathus and Pseudopolygnathus, and were established previously (^{Becker et al., 2013}), the morphological complex group of Protognathodus faunas with atypical diagnostic features could be regarded as Protognathodus-like "protognathoids', and can have morphological features of the late Protognathodus, of Bispathodus and of Gnathodus regarding platform ornamentation and shape of the cup as discussed below. Also, homeomorphic morphological features could be problematic because the early protognathodids can have overlapping ranges with the late Protognathodus (e.g., Protognathodus praedelicatus, see ^{Habibi et al., 2008}; see discussion in ^{Hartenfels et al., 2022}), with Bispathodus (e.g., Bispathodus stabilis,^{Ziegler 1973}, ¹⁹⁷⁴) and with Gnathodus (see below). These atypical faunas ("protognathoids') probably originated during the early phase of the Hangenberg Crisis (Drewer Event) in the praesulcata Zone, with ?Protognathodus meischneri and ?Protognathodus collinsoni, and evolved probably during the anoxic (HBSE) or the regressive phase (HSSE) of the Hangenberg Event, with ?Protognathodus semikockeli and ?Protognathodus kockeli (see Figure 13), and spreaded around the DCB and in the lower Tournaisian. Their occurrence could be related to an opportunistic lifestyle during major environmental changes related to the Hangenberg Event, which can be regarded as "bottleneck" for the evolution of the Idiognathodontidae to which Protognathodus belong. Because protognathoids spread in the kockeli Zone, but mainly in the sulcata/kuehni Zone, or probably even later (Figure 14; see ^{Corradini et al., 2011}; ^{Kaiser et al., 2020}), the extended kockeli Zone (base of the former kockeli Zone to the base of the bransoni Zone) is not applicable due to the spread of these still unknown faunas regarding their stratigraphic and distributional pattern, and also due to their unknown ancestor and descendent relationship.

At Trolp, a stratigraphically leaked younger fauna is recorded by single specimens of Gnathodus at the same stratigraphic level as the early Protognathodus fauna, and leaked Gnathodus were found in the sulcata/kuehni Zone, or in older (praesulcata Zone) levels. Mixed faunas and homeomorphy (see below) within the Idiognathodontidae family (Gnathodus and Protognathodus) could produce a bias of the faunal record in the lower Tournaisian, especially in 1) condensed successions, 2) in stunted or juvenile-dominated assemblages, 3) in Lower Carboniferous, siliciclastic-dominated successions with fossilpoor limestones and scarce conodont faunas, without associated other diagnostic faunas and/or co-occurrence with "homeomorphic'faunas (e.g.,^{Kaiser et al., 2020}; see ^{Königshof et al., 2021} and ^{Parvizi et al., 2021}). At Shahmirzad, (Alborz Mts., northern Iran), Protognathodus kockeli even occurs at same level as Gnathodus cuneiformis and Protognathodus praedelicatus in the upper Tournaisian Lower typicus Zone to anchoralislatus Zone (^{Habibi et al., 2008}, see discussion in ^{Hartenfels et al., 2022}). In contrast, leaked faunas among the late (Siphonodella bransoni, Siphonodella duplicata, Siphonodella quadruplicata) siphonodellids are generally easy to distinguish from the early (Siphonodella praesulcata, Siphonodella sulcata) siphonodellids, as at Trolp in mixed faunal assemblages around the thin-bedded limestone successions (Figure 10a).

^{Lane et al. (1980)} provided an extended discussion of the differences between Gnathodus and Protognathodus and the differences in cup shape and cup terminations between early and late Protognathodus, with diagnoses of both genera. However similar diagnostic features between the early Protognathodus fauna, the late Protognathodus fauna ([Protognathodus cordiformis], Protognathodus praedelicatus, see also discussion in ^{Hartenfels et al., 2022}), and Gnathodus (^{Ziegler, 1973}; ^{Lane et al., 1980}; ^{Dzik, 1997}, ²⁰⁰⁶), and their overlapping ranges in the lower, middle or upper Tournaisian must be considered.

Examples of atypical morphotypes or homeomorphy as discussed as follows are partly described in synonymies for Protognathodus kockeli (^{Hartenfels et al., 2022}). Atypical " Protognathodus' specimens can have a cupmorphology, which is similar to the late protognathodids regarding the termination of the margin and the wideness of the cup, and also confusion with Gnathodus may be related to the platform ornamentation of Protognathodus kockeli, where, for example, nodes as ridgeforming row of "denticles", partly with an adcarinal through, are developed (e.g., ^{Corradini et al., 2011}, plate 1, figure 18; ^{Kaiser et al., 2020}, plate 2, figure 1), or incipient parapet-like morphologies on one side of the platform and a strong asymmetry, e.g., in Sardinia, ^{Corradini et al. (2011, pl. 1}, Figure 15), at Trolp, ^{Kaiser et al. (2020; plate 3, figure 19, 22)}, in Morocco, Lalla Mimouna, ^{Becker et al. (2013}, plate 3, figure 9, with an unusual wide posterior carina, similar to specimen from Oklahoma figured by ^{Over, 1992}, see below), or in China, ^{Wang and Yin (1988},; Figure 15).

Protognathodus specimens from the DCB transition were previously determined as Gnathodus in the Rhenish Massif (e.g., ^{Luppold et al., 1984}, pl. 6, Figure 2) or as Protognathodus praedelicatus in the Carnic Alps or Montagne Noire (^{Flajs and Feist, 1988}; ^{Schönlaub et al., 1988}). Subjective determinations of Protognathodus species from North Africa (^{Korn et al., 2004}), and from China, are evident (^{Wang and Yin, 1988}: pl. 22, Figure 1-19; and compare synonymy list in ^{Corradini et al., 2011}, and discussions therein, p. 25). ^{Over (1992)} described Protognathodus n. sp. from the kockeli Zone (= Upper praesulcata Zone) from North-America, which can be distinguished after the author from Gnathodus by its relatively symmetrical platform and lack of the offset of the outer and inner anterior platform margins. Specimens of Protognathodus kockeli from this region (Oklahoma, Woodford Shale) have a wide posterior carina, similar to that figured specimen from Morocco (see above, ^{Becker et al., 2013}). This species is regarded by ^{Over (1992)} as an intermediate between Protognathodus kockeli and Protognathodus sp. B. ^{Over (1992)} suggested that Protognathodus sp. B may represent the ancestor of Gnathodus punctatus (with FAD in the late Kinder-hookian) which h as a broadly expanded posterior carina. Similar specimens have been illustrated from La Serre, Montagne Noire (^{Flajs and Feist, 1988, pl. 9}, Figure 9, 10), and fused nodes on the platform (plate 9, figure 10) could be transitional to Protognathodus praedelicatus but was determined as Protognathodus kockeli after ^{Corradini et al. (2011)}.

Furthermore, juvenile specimens of Protognathodus meischneri and Bispathodus stabilis are difficult to distinguish (^{Ziegler, 1973}). Both have overlapping ranges in the ?upper/uppermost Famennian and lower Tournaisian, and confusions with stunted or juvenile ?Protognathodus meischneri and Bispathodus stabilis occur at Trolp and elsewhere (^{Kaiser et al., 2020}; see also Corradini et al., 2016). ^{Ziegler et al. (1974)} have demonstrated the ancestry of Protognathodus; Bispathodus stabilis Morphotype 2, which has an expanded basal cavity extending to the posterior tip of the blade (end of the platform), was shown to be the ancestor of the Protognathodus lineage by the authors, illustrated on plate 3, Figure 2. Bispathodus stabilis Morphotype 2 lacks only the wider expansion of the cup to distinguish it from Protognathodus meischneri, and Protognathodus meischneri has a distinctly flat basal cavity.

However, as stated in ^{Kaiser et al., (2020)}, the shape of the cup of adult specimens of ?Protognathodus meischneri at Trolp (Figure 13, narrow shape of cup) can be similar to adult specimens of Bispathodus stabilis stabilis (see ^{Hartenfels, 2011}), and thus confusions in the determinations of both species exist, also in regard to the ornamentation of the platform which can be similar between Protognathodus collinsoni and Bispathodus stabilis bituberculatus and Bispathodus aculeatus, or between Bispathodus stabilis stabilis and Protognathodus meischneri (^{Hartenfels, 2011}; ^{Hartenfels and Becker, 2016}; see ^{Hartenfels, 2011}, plate 31), e.g.,^{Matyja et al., (2021}, Figure 5A d: aff. Protognathodus meischneri);^{Königshof et al., (2021)}; ^{Becker et al., (2013}, plate 3, Figure 5: Protognathodus meischneri [?with initial nodes on the platform]); ^{Matyja et al., (2021}, with a doubtful Protognathodus,Figure 7a, b, c [similar to her Bispathodus aculeatus]);^{Kumpan et al., (2021}, Protognathodus cf. meischneri,Figure 5. 7 [with an initial node on the platform]). The doubtful specimens of Protognathodus meischneri or Protognathodus collinsoni, partly discussed in ^{Hartenfels et al., (2022)}, occur in preextinction levels (praesulcata Zone) or in the Hangenberg Crisis level as single specimens in hemipelagic or neritic settings at La Serre É, Montagne Noire (^{Flajs and Feist, 1988}; ^{Kaiser, 2005}, ²⁰⁰⁹; ^{Feist et al., 2021}), Lesní Lom, Moravian Karst (^{Kumpan et al., 2021}), Mighan and Chelcheli, Alborz Mts., Iran (^{Parvizi et al., 2021}; ^{Bahrami et al., 2021}; ^{Königshof et al., 2021}), Lalla Mimouna, Anti-Atlas, Morocco (^{Becker et al., 2013}), Kowala, Holy Cross Mountains, Poland (^{Matyja et al. 2021}), Grüne Schneid, Carnic Alps (^{Corradini et al., 2017}; ^{Kaiser et al., 2020}). Also, the occurrence of questionable Protognathodus meischneri and Protognathodus collinsoni in pre-Hangenberg Event levels from different regions (e.g. at La Serre E) must be considered in respect to their FAD and ancestor relationship to the new species Protognathodus semikockeli recently established (see also discussion in ^{Hartenfels et al., 2022}).

In summary, beside the diagnostic features of the shape and curvature of the cup of the Protognathodus fauna see Figure 13), morphological complexity of the ornamentation of the platform may resulted in subjective determinations. Therefore, these morphological complexities in shape of the cup and ornamentation of the platform ("protognathoids') should be focused on in future studies, comparable to the "siphonodelloids' which are discriminated from the homeomorphic siphonodellids.

4.2.3 THE DEVONIAN-CARBONIFEROUS BOUNDARY AND CURRENT INDICATORS IN THE SULCATA/KUEHNI ZONE

At the base of the sulcata/kuehni Zone, the 2nd faunal recovery among conodonts (Figure 14) is connected with an increasing abundance of pseudo-polygnathids and siphonodellids (e.g.,^{Kaiser, 2005}; ^{Kaiser et al., 2017}), and with a change from poor, rare and stunted conodont faunas in the ckl/kockeli Zone (1st radiation phase) to more abundant and normal-size faunas indicating the post-extinction/post-crisis level. The 1st and 2nd recovery phases (see Figure 12) each coincide with global post-extinction sea-level rises and deepening episodes, as interpreted, for example, by microfacies at Trolp (^{Kaiser, 2005}) and Grüne Schneid (^{Kaiser, 2007}). At Trolp and Grüne Schneid, it is characterized by more fine-grained micritic limestones - mud- and wackestones - when compared to under- and overlying beds which are pack- and floatstones. The 2nd faunal recovery phase are about time-equivalent to the return to full carbonate depositional conditions and continuing in the lower Tournaisian (middle sulcata/kuehni Zone, Gattendorfia level, 3rd faunal recovery phase). The onset of a distinctive decrease of δ¹³C_carb , values started in the sulcata/kuehni Zone at Trolp with the onset of the 2nd radiation phase, and can be correlated to numerous other settings in European, Asian and North-American regions (see Section 4.1.3).

Between the 1st (base of kockeli Zone) and 2nd recovery phases, a short regressive episode (DCB regression) corresponds to significant survivor extinctions in ammonoids, brachiopods, trilobites and foraminifers at the end of the kockeli Zone to a roughly contemporaneous terrestrial extinction, and to abruptly decreasing carbon isotope values globally recorded (Figure 14; e.g.,^{Kaiser et al., 2006}, ²⁰¹⁶). The level of the DCB regression enables a correlation of pelagic (conodonts, ammonoids) with shallow-water environments (brachiopods, corals, see review in ^{Kaiser et al., 2016}), and can also be correlated into the terrestrial realm by the LN/VI miospore Zone boundary (e.g., ^{Clausen et al., 1994}). The DCB regression is indicated at Trolp by microfacies change from micritic limestones to bioclastic pack- to floatstones reported by ^{Kaiser (2005)} and ^{Bojar et al. (2013}, Figure 3c) although the latter authors assigned this level to the Upper praesulcata Zone based on the uncomplete conodont record as mentioned in Section 4.1.3.

The different morphotypes of Siphonodella sulcata, especially morphotypes 4 and 5, are globally widespread (see Section 4.2.3.1), and Protognathodus kuehni has a widespread geographic distribution which is reported in numerous studies from the sulcata/kuehni Zone (^{Corradini et al., 2011}; Kaiset et al., 2020; see discussion in ^{Hartenfels 2022}). However, this taxon can have, as well as Protognathodus kockeli, a late entry, for example at the GSSP at La Serre É (e.g., ^{Flajs and Feist, 1988}; ^{Feist et al., 2021}). Therefore, the sulcata/kuehni Zone has a high global correlation potential and the FO's of Siphonodella sulcata and Protognathodus kuehni can be correlated to many other regions globally (see ^{Kaiser and Corradini, 2011}; ^{Corradini et al., 2011}; and extended references in ^{Kaiser et al., 2020}; ^{Königshof et al., 2021}; ^{Qie et al., 2021}; ^{Corradini et al., 2021}; ^{Spalletta et al., 2021}; ^{Kumpan et al., 2021}; ^{Matyja et al., 2021}; ^{Becker et al., 2021}; ^{Feist et al., 2021}; ^{Komatsu et al., 2014}). While the extended kockeli Zone leaves the event-related successions and the DCB undivided, as recently used for example at Grüne Schneid, La Serre E, and Milles (^{Feist et al., 2021}, ^{Spalletta et al., 2021}; ^{Aretz et al., 2021}; Figure 12), a precise conodont biozonation at the DCB can be applied by the use of Siphonodella sulcata M5 and Protognathodus kuehni.

At Trolp, the typical morphotype of Protognathodus kuehni and Siphonodella sulcata M5 first occur at same stratigraphic levels (Figure 10a, 12). Also, Siphonodella sulcata ?M4 was figured in ^{Nössing (1975)} in the basal sulcata/kuehni Zone. The sulcata/kuehni Zone can thus be established in the Graz Paleozoic, and at correlative levels in the Carnic Alps and Moravian Karst, based on the cooccurrence of both Siphonodella sulcata M5 and Protognathodus kuehni (Figure 12; ^{Kalvoda et al., 2015}; ^{Kaiser et al., 2020}; ^{Kumpan et al., 2021}). Sampling of continuous limestone successions there supports the coeval FAD of both taxa at the base of the Tournaisian. After ^{Kaiser et al. (2020)}, this supports the use of a joint biozone in eventrelated successions, because their cooccurrences are evidenced from these three different regions; the utility of a joint biozone enhances the possibility of using either the biostratigraphic index Siphonodella sulcata or Protognathodus kuehni, if one or the other h as a late entry or is as yet unrecorded in a region due to facies-controlled rare or diachronous occurrences (e.g.,^{Kaiser et al., 2009}; ^{Corradini et al., 2011}; ^{Spalletta et al., 2017}).

4.2.3.1. CONSIDERATIONS OF SIPHONODELLA

Within siphonodellids, the abundance of intermediate forms attests to the completeness of the stratigraphic record as at La Serre É (Figure 12; between Siphonodella praesulcata, Siphonodella sulcata M4 and M5, and advanced siphonodellids); the differences between Siphonodella praesulcata and Siphonodella sulcata are stated in ^{Sandberg et al. (1972)}. However, the occurrence of intermediate forms may have resulted in subjective determination of the first occurrence of Siphonodella sulcata (e.g.,^{Ziegler and Sandberg, 1996}; ^{Kaiser, 2009}; ^{Becker et al., 2016a}), but Siphonodella-morphotype groups can be clearly distinguished and recorded the spectrum of morphotypes and intermediates, which spread in the lower Tournaisian (^{Kaiser and Corradini, 2011}). These groups are used to distinguish between Siphonodella praesulcata (M2, M3) and Siphonodella sulcata (M1, M4-M7), and also between early (M4, M5) and advanced forms (M1, M6, M7) of Siphonodella sulcata, the latter are recorded especially in younger Tournaisian successions at La Serre É (^{Kaiser and Corradini, 2011}). The morphotype groups are based mainly on the curvature of the carina, a feature which had been used for the current GSSP level at La Serre É (^{Flajs and Feist, 1988}, see also ^{Sandberg et al., 1972}), as well as shape of the platform, a pseudokeel which contains a pit (basal cavity), and the ornamentation. Siphonodella sulcata Morphotype 4 already has slight constrictions anteriorly but is otherwise similar to Siphonodella sulcata M5. The Siphonodella sulcata M1, M6 and M7 are representatives of advanced forms, and are intermediates between Siphonodella sulcata and Siphonodella bransoni (former Siphonodella duplicata M1) because they are strongly bended and have constrictions anteriorly.

Siphonodella sulcata M4 and M5 are easy to recognize, most widespread, and common in North America, Asia, and Europe (Figure 12, ^{Kaiser and Corradini, 2011}, figure 7; see ^{Kaiser et al. 2020}). The globally most widespread Siphonodella sulcata M5 is the earliest morphotype which occur at the base of the Tournaisian, for example at Trolp, Grüne Schneid (^{Kaiser et al., 2020}), La Serre É (^{Kaiser, 2009}; ^{Kaiser and Corradini, 2011}; ^{Feist et al., 2021}), Lalla Mimouna (Anti-Atlas, Morocco, ^{Becker et al., 2013}), Lesni Lom (Moravian Karst, Czech Republic; ^{Kalvoda et al., 2015}), Kowala (together with Protognathodus kuehni and Polygnathus purus subplanus, Holy Cross Mountains, Poland; ^{Malec, 2014}; ^{Matyja et al., 2021}), Borkewehr (Rhenish Massif, ^{Hartenfels et al., 2022}). Siphonodella sulcata M4 was found in lower Tournaisian successions, for example at Bou Tlidat (Maider, Anti-Atlas, Morocco; collection S.I. Kaiser; see ^{Kaiser et al., 2011}; ^{Becker and Kaiser, 2023}), Mighan (Alborz Mts. northern Iran, ^{Parvizi et al., 2021}) and at Shahmirzad (Alborz Mts.; ^{Habibi et al., 2008}). Siphonodella sulcata M4 evolved from Siphonodella sulcata M5 based on the record at La Serre É, and represents most probably the marker for the 3rd radiation phase in the sulcata/kuehni Zone at the Gattendorfia level (sulcata Event). Siphonodella bransoni is also recorded from successions at M'Karig, Tafilalt, Anti-Atlas (collection S.I. Kaiser; see ^{Kaiser et al., 2011}; ^{Becker and Kaiser, 2023}). For the occurrence of Siphonodella praesulcata in neritic settings in Morocco, Iran, China, Vietnam, see ^{Kaiser (2005)}, ^{Kaiser et al. (2011)}, ^{Becker et al. (2013)}, ^{Komatsu et al. (2014)}, ^{Parvizi et al. (2021)}, ^{Bahrami et al. (2021)}, ^{Qie et al. (2021)}. In shallow-water settings at Gedongguan, Muhua, Nanbiancun and Dapoushang (South China), Siphonodella sulcata occurs, also in association with Protognathodus kuehni, at the base of the Tournaisian (^{Qie et al., 2021}), and Siphonodella sulcata is also reported from shallow-water settings for example from northeastern Vietnam (^{Komatsu et al., 2014}). In summary, siphonodellids also occur in neritic successions, and the sulcata/kuehni Zone can be established in pelagic, hemipelagic and neritic settings based on Siphonodella sulcata and/or Protognathodus kuehni. However, an evaluation of various records from different paleogeographical settings of the earliest Morphotype 5 and Morphotype 4 of Siphonodella sulcata at the base of the Tournaisian is needed (e.g., ^{Becker et al., 1984}; ^{Qie et al., 2021}; ^{Komatsu et al., 2014}; ^{Matyja et al., 2021}, ^{Hartenfels et al., 2022}; see also ^{Aretz et al., 2021}, and references therein).

Siphonodella sulcata M4 and M5 can be clearly distinguished from Siphonodella praesulcata, from the late siphonodellids (see Section 4.2.2.1), and from Siphonodella-like "siphonodelloids'; the latter have overlapping ranges with Siphonodella praesulcata mainly in the praesulcata Zone, as documented e.g., in ^{Becker et al. (2013)}. The "siphonodelloids' originated and evolved in pre-Hangenberg event levels in the expansa Zone probably connected with enhanced C_org burial phases and/or shallowing phases during the Epinette and Etreoungt Events (for environmental changes during the events see ^{Kaiser et al., 2008}), and spread in the praesulcata Zone; its occurrence is related to biotic opportunism comparable with the spread of "protognathoids' during the Hangenberg Event (Figure 14). "Siphonodelloids' are most widespread in hemipelagic or shallow-water settings, for example in Morocco (e.g.,^{Becker et al., 2013}), Iran (e.g.,^{Parvizi et al., 2021}), and probably at La Serre É (Figure 14).