Conodonts: The new sampling confirmed not only the previous conodont ranges but was also very successful because of the first record of L. b. gracilis in the Požár 3 section. The taxon L. b. gracilis with its characteristic development of sharp connecting ridges on the inner lateral process (see Fig. 5A–D) is considered as the oldest one within the lower Emsian part of the Latericriodus clade and can be well correlated around peri-Gondwana (see the discussion in Weinerová et al. 2024). As already mentioned, the entry of L. b. gracilis is 0.8 m below the base of the BGE interval, it means much higher than in the Mramorka section, where the taxon enters 145 cm below the base of the BGE interval (Weinerová et al. 2024: Fig. 4). This means that Mramorka, the section with almost identical lithology and similar accumulation rate in the upper parts of the Praha Formation as in the Požár 3 section, remains the best candidate for the basal Emsian GSSP because of the lowermost occurrence of this prospective GSSP-defining conodont marker.
The data also confirm extreme scarcity of polygnathids, as no single Polygnathus has been found in the Požár 3 section so far. Accordingly, the only polygnathid specimen recorded in the Požáry quarries came from the early part of the Praha Formation from the former Požár 2 Quarry identified as Polygnathus pireneae Boersma, 1973 (Slavík 2004a). There is, however, a possibility of occurrence of younger polygnathids as those found by Kalvoda in the Mramorka section (1995: Pl. 2, Fig. 9) and Schönlaub in Chlupáč et al. (1986), above the BGE beds. But, the real entry of the Polygnathus excavatus excavatus group can be expected lower, even below the Moroccan equivalent of the BGE, the atopus Event in the early gracilis Biozone (see Aboussalam et al. 2015: Fig. 31), but due to the enormous scarcity it is difficult to locate it in spite of many resamplings of the early Devonian sections in the Prague Synform. Therefore, the taxa of the genus Polygnathus (= Eolinguipolygnathus Bardashev, Weddige and Ziegler, 2002) are not reliable for an identification of the boundary in the Prague Synform.
Lithology and ichnofabric
The Mramorka and Požár 3 sections are similar in facies development of the Praha Formation (cf. Chlupáč 1957; Koptíková et al. 2010a, b). Accordingly, microfacies obtained from closely studied interval around the BGE beds at the Mramorka section (Weinerová et al. 2024) and those described in this study from the Požár 3 section are also very close. They are represented by wackestones/packstones to floatstones with allochems dominated by shells of planktonic dacryoconarids, but they contain also microproblematicum Globochaete Lombard, 1945, and remains of nekton (e.g. nautiloids) and heterotrophic benthos. The high content of fine-grained matrix and allochem composition are suggestive of a deep-subtidal environment. Common bioerosion of some bioclasts (molluscs, trilobites, crinoids) could possibly indicate slow sedimentation rates (e.g. Lindsröm 1979). High bioclast fragmentation and telescoping (cone-in-cone stacking) of dacryoconarid shells might indicate sediment transport. These limestones can represent turbidites originating from turbulent gravity surges as previously suggested by Hladil et al. (1996, 2014), but other transport mechanisms cannot be completely excluded. Distribution of bioclasts within the sediment is at least in some cases influenced by common bioturbation. As in the Mramorka section, Balanoglossites burrows filled by by Chondrites-type burrows which underwent dolomitization were observed.
Gamma-ray spectrometry (GRS) and magnetic susceptibility (MS): GRS and MS logs through the whole Praha Formation at the Požár 3 section were published by Koptíková et al. (2010a, b) and Hladil et al. (2010, 2011). Koptíková et al. (2010a) interpreted increasing/decreasing of K and Th contents in terms of dilution of fine-grained terrigenous material in carbonate due to a variation in carbonate productivity and/or siliciclastic input during transgressions/regressions, whereas Th/U ratios were used as a redox proxy. MS signal according to Koptíková et al. (2010b) reflect fine-grained siliciclastic admixture in carbonates, but is further modulated by Fe-bearing diagenetic minerals. The sequence stratigraphic interpretation of Devonian sequences in the Prague Basin was published by Bábek et al. (2018a: Fig. 13). Considering therein studied part of the Požár 3 section K, Th, CGR and MS values below the BGE interval are slightly higher than those above this interval. Similarly, to the Mramorka section (see Weinerová et al. 2024), the part of the section below the BGE interval can be ascribed to culminating transgression, whereas the part above the BGE interval to highstand regression. BGE interval level in the Prague Synform was supposed to be related to maximum flooding by Bábek et al. (2018a). Accordingly, the slightly elevated U/Th values observed within the BGE interval at the Požár 3 section can be ascribed to less-oxic condions during the maximum flooding.
Stable isotopes
Therein obtained δ18O and δ13C values of whole-rock samples correspond mainly to deep burial-marine calcite mixes, but are more close to marine-calcite (Fig. 7, cf. with Hasiuk et al. 2016). The δ18O values ranging ca from − 4 to -2% are comparable with values of whole-rock limestone samples and brachiopod shells known from most other coeval localities in the Prague Synform such as the Na Branžovech, Stydlé vody and Černá rokle sections (ca -6 to -1‰, see Hladíková et al. 1997, 2000; Gessa and Lécuyer 1998; Weinerová et al. 2020). In this respect, the Požár 3 section provided less altered δ18O record in comparison to the Mramorka section, as in the latter, the δ18O and δ13C values are more close to the deep-burial calcite, and the δ18O values below and within the BGE interval are exceptionally low (ca -10 to -4‰), see Weinerová et al. (2024).
The δ13C data from whole Praha Formation at the Požár 3 section published by Buggisch and Mann (2004) show two positive peaks (Pragian and lowermost Emsian), which can be recognized also in other sections in the Prague Synform (Weinerová et al. 2020: Fig. 9). New data focusing in detail on interval around the BGE beds reveal that δ13C values above the BGE beds are higher than those obtained from and below the BGE beds. This feature was recognized also at the Mramorka section (Weinerová et al. 2024).
Multi-element geochemistry (INAA data). Similar to the Mramorka section, the PCA on clr transformed INAA data from the studied part of the Požár 3 section allow for discrimination of main and trace elements into three main groups: elements bound prevalently in carbonates (Ca, Mg, Sr, Mn and Br), elements concentrated mainly in terrigenous input (Al, K, Th, Ti, Rb, Cr, Hf, Cs, Na, Fe, V, Sc) and elements which were sensitive to palaeoredox and/or palaeoproductivity changes (U, Co, As and to less extent also Zn) (see e.g. Veizer 1983; Tribovillard et al. 2006; Smrzka et al. 2019; Carter et al. 2020).
In therein studied part of the Požár 3 section, the elements concentrated mainly in terrigenous input show generally a decreasing trend – limestones below the BGE interval have slightly higher concentrations of these elements than limestones within and above this interval. CIW´, Na/Ti and K/Na ratios can be used as proxy of chemical weathering (Nesbitt et al. 1996; Cullers et al. 2000; Song et al. 2013; Sayem et al. 2018). In the Mramorka section, a decrease in CIW' and K/Na and an increase in Na/Ti above the BGE beds were observed, which may possibly indicate a decrease in the intensity of chemical weathering or a change in the detrital material provenance (Weinerová et al. 2024). Nevertheless, the data from the Požár 3 section does not confirm this trend.
Elements bound preferentially in carbonates show higher concentrations in higher parts of the section, i.e. within and/or above the BGE interval. Despite the serious pitfalls (see Bausch 1968; Tucker 1986; Ahm et al. 2018), e.g. Mn possibly partly bound in Mn oxides (which cannot be excluded especially in the BGE interval in this study), the Mn/Sr ratios are commonly used as a criterion for screening diagenetic alteration in carbonates (Brand and Veizer 1980; Ganai et al. 2018; Higgins et al. 2018). Samples from the Požár 3 section show values below the threshold of 2 and no correlation between Mn and Sr (r = 0.04), indicating that the limestones have not experienced any significant diagenetic changes according to this criterion.
Enrichment factors of some elements used as palaeoredox/palaeoproductivity proxies (e.g. UEF, AsEF, BaEF, ZnEF, CoEF) are elevated in the BGE interval suggesting decrease in oxygenation/higher palaeoproductivity, which is in accord with maximum flooding supposed by previous authors (Bábek et al. 2018a). Nevertheless VEF does not show this trend and a very strong positive correlation between V and Al (r = 0.93) suggests that V is bound mainly in material of terrigenous origin in the Požár 3 section. For comparison, the BGE interval at the Mramorka section shows elevated VEF (and also ScEF) but does not show elevated values of CoEF, see Weinerová et al. (2024). The information obtained from enrichchment factors is needed to take with a caution with respect to low Al contents (see Tribovillard et al. 2006; Kumpan et al. 2019)
Multielement geochemistry (EDXRF data). The less accurate but more numerous EDXRF measurements complement the INAA results. The moderate correlation between Pb and Al (r = 0.52), indicate that the distribution of this element might be controlled by other factors besides the siliciclastic input – Pb was used as a palaeoproductivity or redox proxy in some studies (e.g. Kumpan et al. 2019; Rakociński et al. 2021, 2023). At the Požár 3 section a higher PbEF values can be observed within and above the BGE interval, i.e. in sediments which are supposed to represent maximum flooding and highstand conditions. In the Mramorka section, higher PEF values were observed above the BGE beds (Weinerová et al. 2024).
Rare earth elements (INAA data). REE from this locality were already studied by Koptíková et al. (2010b), who focused on the whole Požár 3 section (Lochkov, Praha and Zlíchov formations). These authors found a similarity between REE distribution pattern of whole-rock limestone samples and REE distribution pattern of aeolian material published by Nozaki (2001), suggesting that most impurities are of aeolian, not fluvial origin.
∑REE from therein studied part of the Požár 3 section show a strong positive correlation (r = 0.67) with and Al, which confirms that REE concentrations are considerably controlled by the quantity and character of non-carbonate (terrigenous) material. The Mramorka section shows stronger positive correlation between ∑REE and Al below the BGE interval (r = 0.97) than within + above the BGE interval (r = 0.48) suggesting other possible sources of REE within the BGE and its close overburden (Weinerová et al. 2024), while the Požár 3 section shows in both these parts very strong positive correlation between ∑REE and Al (r = 0.86 and 0.95, respectively).
Almost all samples show MREE bulge and their REEN distribution pattern can be compared to REEN distribution pattern of terrigenous sources such as SSWR (Suspended Sediment World River, Viers et al. 2009), WRAC (World River Average Clay, Bayon et al. 2015), marine atmospheric dust (Greaves et al. 1994), or to anoxic pore waters (Haley et al. 2004); see Fig. 11. Nevertheless, the best match was found with the marine atmospheric dust (Greaves et al. 1994), which is in accord with results of Koptíková et al. (2010b). Some samples from the BGE interval and its close overburden show slightly higher HREEN values and their REEN distribution pattern can be partly correlated also with REEN pattern of oceanic mean dissolved concentrations (Nozaki 2001), carbonates (microbialites, Webb and Kamber 2000), diagenetic ferromanganese crusts/nodules (Bau et al. 2014) and oxic/suboxic pore waters (Haley et al. 2004). This feature is reminiscent of the situation on the Mramorka section. However, in the Mramorka section, most samples from the BGE interval and its close overburden show higher HREEN values and a linear REEN distribution pattern, which is closer to the distribution patterns of oceanic mean dissolved concentrations, microbialites, diagenetic ferromanganese crusts/nodules, and oxic/suboxic pore waters than to aeolian distribution pattern.