5.a. Palynology: palynofloral turnover and palaeoclimatic background

The T–J boundary is historically the least well studied of the all faunal mass extinction events and has been particularly lacking data with regard to plant palaeoecology. However, available information hints at a major palaeoecological disturbance of plant communities. Dating of the extinction event was and still is a much debated issue (e.g. Fowell & Olsen, Reference Fowell and Olsen1993; van Veen, Reference Van Veen1995; Olsen et al. Reference Olsen, Kent, Sues, Koeberl, Huber, Montanari, Rainforth, Fowell, Szajna and Hartline2002; Lucas & Tanner, Reference Lucas and Tanner2007). In his seminal work, Harris (e.g. Harris, Reference Harris1937) described fossil flora of the Scoresby Sund (East Greenland), and he distinguished two palynofloral assemblages: a Rhaetian Lepidopteris Zone and a Hettangian Thaumatopteris Zone. Studies of plant taxonomic ranges have revealed that up to 90% of species became locally to regionally extinct across North America (Cornet & Olsen, Reference Cornet, Olsen and Weber1985; Fowell & Olsen, Reference Fowell and Olsen1993) and Greenland (Pedersen & Lund, Reference Pedersen and Lund1980; McElwain, Beerling & Woodward, Reference McElwain, Beerling and Woodward1999; McElwain et al. Reference McElwain, Popa, Hesselbo, Haworth and Surlyk2007). In Greenland, considerable changes in the dominance and diversity structure of Triassic forests have been identified before the T–J boundary. Triassic communities of Podozamites, a broad-leafed conifer, and bennetites (Pterophyllum and Anomozamites) were replaced by lower diversity and less even (i.e. more dominant) forests, overwhelmingly dominated by taxa that had been relatively minor components of Late Triassic forests: Czekanowskia, Sphenobaiera and an osmundaceous fern (Todites) (McElwain et al. Reference McElwain, Popa, Hesselbo, Haworth and Surlyk2007; McElwain, Wagner & Hesselbo, Reference McElwain, Wagner and Hesselbo2009). Importantly, a climate-driven shift from a prevalence of broad-leaved taxa to a predominantly narrow-leaved assemblage contributed to increased fire activity at the T–J boundary in Greenland (Belcher et al. Reference Belcher, Mander, Rein, Jervis, Haworth, Hesselbo, Glasspool and McElwain2010a). However, estimates of diversity loss based on macrofossils are typically much higher than estimates of diversity loss based on miospores (Mander, Kürschner & McElwain, Reference Mander, Kürschner and McElwain2010). The authors explained that conflicting records of diversity loss obtained from plant macrofossils and sporomorphs are caused by the absence of reproductively specialized plants, including cycads, bennetites and the seed-fern Lepidopteris, from the palynofloral record. Moreover, current reports from Europe and the Tethyan domain (Kürschner & Herngreen, Reference Kürschner, Herngreen and Lucas2010; Cirilli, Reference Cirilli and Lucas2010) show that palynofloral composition between the Late Triassic and Hettangian was gradual and without abrupt changes and consequently claimed that the end-Triassic biotic crisis appears to have little affected palynofloral species diversity, at least in Europe. On the other hand, in North America high-diversity pollen assemblages comprising monosulcates and monosaccates give way to lower-diversity assemblages dominated by Classopollis, a pollen type normally associated with hot and/or arid climate conditions, and palynofloral diversity loss is estimated at about 60% (Fowell & Olsen, Reference Fowell and Olsen1993; Lucas & Tanner, Reference Lucas and Tanner2007). These low-diversity pollen assemblages are confined to within 21000 years before the T–J boundary in eastern North America as defined by a fern spike (Fowell & Olsen, Reference Fowell and Olsen1993) and negative carbon isotope excursion. Recently, the fern spike has been correlated with the ‘initial’ carbon isotope excursion and end-Triassic extinction (Whiteside et al. Reference Whiteside, Olsen, Eglinton, Brookfield and Sambrotto2010).

In Kamień Pomorski the palynofloral turnover coincides with the sequence boundary at 678.4 m. Twenty new taxa appear just above this boundary and within 2 m above it. Some typically Rhaetian spores pass into the strata immediately above the sequence boundary, but they show rapid decrease in frequency (Cingulizonates rhaeticus and Limbosporites lundblandii) and quickly disappear. Importantly, the index pollen grain Cerebropollenites thiergartii appears just at the sequence boundary. Several characteristic Jurassic taxa (Concavisporites intrastriatus, Concavisporites umbonatus, Conbaculatisporites mesozoicus, Dictyophyllidites mortoni, Zebrasporites interscriptus and Cerebropollenites thiergartii) show continuity throughout the rest of the Hettangian profile studied herein. Chordasporites platysaccus, Pinuspollenites minimus and Vitreisporites pallidus appear already in the Rhaetian strata and continue through the Jurassic section. As was previously mentioned, a hiatus at the sequence boundary could exist, which would make the transition perhaps appear more abrupt than it actually is.

The Rhaetian assemblage from Pomerania (which is referred to here as the Cingulizonathes rhaeticus–Limbosporites lundblandii association; Fig. 5) can be compared to other biostratigraphical/palynological zones in Europe, both in Tethyan and epicontinental domains (Fig. 5; Table 2; see Bonis, Kürschner & Krystyn, Reference Bonis, Kürschner and Krystyn2009; Lund, Reference Lund1977; Pedersen & Lund, Reference Pedersen and Lund1980; Kürschner, Bonis & Krystyn, Reference Kürschner, Bonis and Krystyn2007). Also in the previous works from Poland (Orłowska-Zwolińska, Reference Orłowska-Zwolińska1983, Reference Orłowska-Zwolińska1985; Marcinkiewicz & Orłowska-Zwolińska, Reference Marcinkiewicz and Orłowska-Zwolińska1994) Limbosporites lundblandii (along with Cingulizonathes rhaeticus and Semiretisporis gothae) are regarded as the most common and characteristic Rhaetian forms in Poland.

Table 2. Occurrences of miospores from Kamień Pomorski in other regions of Europe

Note clear difference with Asturias, Spain, pointing to provincialism between North-Central Europe and the Mediterranean area.

The narrow interval between 677 and 680.9 m (Fig. 3) could represent the transitional Ricciisporites–Polypodiisporites (Lund, Reference Lund1977) or Trachysporites–Porcellispora Zone (Kürschner, Bonis & Krystyn, Reference Kürschner, Bonis and Krystyn2007), although in Pomerania the transitional assemblage cannot be clearly distinguished. Only Ricciisporites tuberculatus and Ovalipollis ovalis occur exclusively within this transitional zone, at the same time crossing the inferred Rhaetian–Hettangian boundary.

The Hettangian assemblage in Pomerania is referred to here as the Conbaculatisporites mesozoicus–Dictyophyllidites mortoni–Cerebropollenites thiergartii association (Fig. 5). All three miospore taxa characterizing Hettangian strata in Pomerania are widespread and characteristic forms, known from most of the important T–J profiles (Lund, Reference Lund1977; Barrón et al. Reference Barrón, Gómez, Goy and Pieren2006; Kürschner, Bonis & Krystyn, Reference Kürschner, Bonis and Krystyn2007; Bonis, Kürschner & Krystyn, Reference Bonis, Kürschner and Krystyn2009 and others; Fig. 5; Table 2). Dictyophyllidites mortoni, Lycopodiumsporites semimuris and Trachysporites fuscus were listed by Fijałkowska (Reference Fijałkowska1989) as characteristic Hettangian forms from central Poland. The two cored sections, 635–642.1 m and 649.8–653.4 m, show continuity of the Jurassic palynological assemblage (eight taxa continue), although 20 new taxa are reported from these two sections. However, these new taxa do not represent any new groups of plants at a higher taxonomic level, and are still represented by Lycopodiacea (Uvaesporites, Lycopodiacidites, Lycopodiumsporites), Pteridophyta (ferns Apiculatisporites, Baculatisporites, Cibotiumsporites, Concavisporites, Cosmosporites, Ischyosporites, Lycopodiacidites, Polycingulatisporites, Stereisporites, Trachysporites, Zebrasporites) plus Bryophyta (rare Stereisporites) as chief parent plants for spores, and Coniferophyta (Pinaceae, Podocarpaceae, Caytoniales) representing parent plants of the pollen grains (Alisporites, Microcachryidites, Quadraeculina, Vitreisporites). As mentioned above, palynofacies and miospore spectra in these two sections from Kamień Pomorski are chiefly dictated by palaeoenvironmental factors, i.e. shifting nearshore-marginal marine depositional environments. Palynofacies thus reflect flooding (transgressive)-progradation cyclicity (Fig. 3): flooding events result in relatively poorer palynomacerals with less abundant miospores and more abundant Acritarcha (interval 649.8–653.4 m), while progradational deltaic–barrier/lagoonal facies (representing delta plain–lagoon/marsh environments) show rich palynomacerals with abundant miospores, particularly spores produced by local hydrophilic plants (interval 637.2–642.4 m). Moreover, some minor fluctuations of palynofacies from lagoonal/interdistributary–delta plain environments, reflecting probable seasonal changes, have been described from the section between 637.2 and 642.4 m depth (Pieńkowski, Reference Pieńkowski2004; Pieńkowski & Waksmundzka, Reference Pieńkowski and Waksmundzka2009).

Comparison of T–J spores described in the present paper show a significant number of common taxa with the adjacent sedimentary basins and the Northern Tethys domain (Table 2). In contrast, there are only six taxa in common with the profiles in Asturias, Spain (of 43 listed there; Barrón et al. Reference Barrón, Gómez, Goy and Pieren2006) and very few (4–6) with the British profiles (Warrington & Harland, Reference Warrington and Harland1975; Warrington, Cope & Ivimey-Cook, Reference Warrington, Cope and Ivimey-Cook1994), but only 20 to 33 miospore taxa were identified in the British profiles at St Audrie's and Larne. Even fewer taxa occur in common with the North American sections (Cornet & Olsen, Reference Cornet, Olsen and Weber1985; Fowell & Olsen, Reference Fowell and Olsen1993; Cirilli et al. Reference Cirilli, Marzoli, Tanner, Bertrand, Buratti, Jourdan, Bellieni, Kontak and Renne2009). Absence of the otherwise widespread thermophilic pollen Corollina sp. (Classopollis sp.) is notable; these pollen were noted in central Poland, some 700 km SE from Kamień Pomorski (Fijałkowska, Reference Fijałkowska1992; Marcinkiewicz & Orłowska-Zwolinska, Reference Fowell and Olsen1994; Ziaja, Reference Ziaja2006). These pollen show very low frequencies in some other regions (i.e. East Greenland), which was attributed to palaeogeographic factors (Pedersen & Lund, Reference Pedersen and Lund1980).

Comparison with other profiles also shows that a few stratigraphically indicative forms found in Kamień Pomorski (such as spores Cingulizonathes rhaeticus, Conbaculatisporites mesozoicus, Limbosporites lundblandii, Ricciisporites tuberculatus, Deltoidospora toralis, Trachysporites fuscus and pollen grains Cerebropollenites thiergartii, Minutosaccus potoniei, Pinuspollenites minimus) are more cosmopolitan than other miospores (Table 2). Collectively, these facts suggest some provinciality of the T–J palynofloral assemblages.

Perhaps, this provinciality and changing local climatic and palaeoenvironmental conditions could provide explanations for the conflicting records of T–J palynofloral turnover. Mander, Kürschner & McElwain (Reference Mander, Kürschner and McElwain2010) and Kürschner & Herngreen (Reference Kürschner, Herngreen and Lucas2010) played down the scale of this turnover, pointing to emigration and/or extirpation of taxa rather than immigration and/or origination of taxa. According to these authors, the end-Triassic biotic crisis appears to have little affected palynofloral species diversity in Europe and Greenland. However, similarly to Fowell & Olsen (Reference Fowell and Olsen1993), we report quite a marked palynofloral turnover at the T–J boundary. Influence of rather minor erosion and a potential missing section at the sequence boundary in Kamień Pomorski does not provide a sufficient explanation. Perhaps, differences between T–J palynofloral assemblages could be regionally enhanced (or the opposite, masked) by changing climatic or taphonomic conditions. This issue needs further studies.

Interestingly, the palynofloral turnover, which coincides with the sequence boundary at 678.4 m, was not, at least not directly, related with crucial change in humidity of the climate: the climate was generally humid throughout deposition of the whole 677–680.9 m interval, which is proven by the presence of rich drifted floral remains, coalified plant roots and a huge amount of spores produced by Lycopodiacea (Selaginellaceae: Limbosporites, Lycopodiumsporites) and Pteridophyta (ferns Cyatheaceae/Dicksoniaceae: Cibotiumsporites, Concavisporites; Dipteridaceae: Conbaculatisporites, Deltoidospora; Matoniacea: Dictyophyllidites; Osmundacea: Baculatisporites; Schizaceae: Duplexisporites). On the other hand, the presence of Coniferophyta (Pinaceae, Podocarpaceae) is proven by the presence of pollen grains such as Ovalipollis potovalis, Quadraeculina sp. and Platysaccus niger below the sequence boundary (Rhaetian section) and Cerebropollenites thiergartii and Taeniasporites sp. above this boundary (Hettangian section). If a rapid increase in humidity was not a direct factor in this marked turnover, then there must have been other environmental factors responsible for the palynofloral change in our material. Climatic changes, including a general increase in humidity, commenced above the first sequence boundary at 698 m. Judging from changes in leaf morphology and decreases in stomatal indices, McElwain, Beerling & Woodward (Reference McElwain, Beerling and Woodward1999) postulated a fourfold increase in atmospheric CO₂ resulting in global warming and sudden loss of Late Triassic biodiversity, for example in East Greenland, a climate-driven shift from broad-leaved to narrow-leaved taxa and resulting fire activity at the T–J boundary (McElwain, Wagner & Hesselbo, Reference McElwain, Wagner and Hesselbo2009; Belcher et al. Reference Belcher, Mander, Rein, Jervis, Haworth, Hesselbo, Glasspool and McElwain2010a). Similarly, increased content of other greenhouse gases such as methane (Hesselbo et al. Reference Hesselbo, Robinson, Surlyk and Piasecki2002) may have contributed to the climate warming. On the other hand, Hubbard & Boulter (Reference Hubbard and Boulter2000) argued for a dramatic series of global temperature oscillations, including at least one colder/drier period at or just above the T–J boundary. Schoene et al. (Reference Schoene, Guex, Bartolini, Schaltegger and Blackburn2010) postulated global cooling and glaciations associated with the end-Triassic extinction period. These contradicting results suggest that the contrasts of Late Triassic climates may have been dramatic enough to trigger radical biological and ecological changes (Hubbard & Boulter, Reference Hubbard and Boulter2000; van de Schootbrugge et al. Reference Van de Schootbrugge, Quan, Lindström, Püttmann, Heunisch, Pross, Fiebig, Petschick, Röhling, Richoz, Rosenthal and Falkowski2009). Indeed, dramatic release of SO₂ and sulphate aerosols (causing climate cooling), following release of CO₂ and methane (causing climate warming), with additional release of toxic pollutants such as PAHs, could have most likely been the final blow to the highly stressed end-Triassic ecosystem, as was postulated by van de Schootbrugge et al. (Reference Van de Schootbrugge, Quan, Lindström, Püttmann, Heunisch, Pross, Fiebig, Petschick, Röhling, Richoz, Rosenthal and Falkowski2009) and Deenen et al. (Reference Deenen, Ruhl, Bonis, Krijgsman, Kürschner, Reitsma and Van Bergen2010). The spore/pollen ratio in the critical boundary section between 677 and 680.9 m (Fig. 3) shows marked fluctuations expressed in palynofacies (Fig. 3), which suggest rapid climate (temperature) changes just below and at the T–J transition. Additionally, darkening of spores observed in this interval, particularly in three samples (Figs 3, 8), could have been caused by soil acidification from sulphuric acid rains during CAMP eruptions (van de Schootbrugge et al. Reference Van de Schootbrugge, Quan, Lindström, Püttmann, Heunisch, Pross, Fiebig, Petschick, Röhling, Richoz, Rosenthal and Falkowski2009). On the other hand, contrasting miospore colours in one sample, named by Pieńkowski & Waksmundzka (Reference Pieńkowski2009) as palynofacies inversion no. 1, can be attributed to the early burial setting: miospores showing darker colours temporarily went through an early diagenetic cycle, commonly in the swamp/marshy environment, and after becoming darker they were redeposited again to be finally incorporated into the sediment. At the same time, the ‘background’ (lighter) miospores were delivered to the sediment directly from parent plants.

The presence of charcoal and the organic geochemistry indicate that despite the humid climate, forest fires occurred on the surrounding lands (Pieńkowski & Waksmundzka, Reference Pieńkowski and Waksmundzka2009; Marynowski & Simoneit, Reference Marynowski and Simoneit2009). A dramatic rise in fire activity has been recently attributed to the dominance of narrow-leaved plants coupled with increased lightening strikes in the much warmer climate at the beginning of the Jurassic period (Belcher et al. Reference Belcher, Mander, Rein, Jervis, Haworth, Hesselbo, Glasspool and McElwain2010a). Noteworthy is that concentrations of atmospheric oxygen reconstructed from recent experimental burn studies using the GEOCARBSULF (Belcher & McElwain, Reference Belcher and McElwain2008) and sedimentary charcoal concentration (Glasspool & Scott, Reference Glasspool and Scott2010) challenged the previous model estimates of a very low concentration of atmospheric O₂, ~ 12–13%, during Late Triassic and Early Jurassic time (Berner, Reference Berner2006; Algeo & Ingall, Reference Algeo and Ingall2007). The new estimates show that the lower O₂ limit for combustion should be increased from 12% to at least 15% (Belcher & McElwain, Reference Belcher and McElwain2008) or 16% (Belcher et al. Reference Belcher, Yearsley, Hadden, Mcelwain and Rein2010b), and based on concentration of charcoal, O₂ atmospheric concentration was much higher, about 25% (Glasspool & Scott, Reference Glasspool and Scott2010). Concerning charcoal concentration, one should take notice that charcoal produced by wildfires was widely redeposited (owing to its resistance to biogenic degradation and buoyancy) and commonly shows an elevated concentration (in comparison to other palynomacerals) in high-energy environments and younger sediments. Reworking and redeposition of charcoal in high-energy environments occurred not only in alluvial environments (Marynowski & Simoneit, Reference Marynowski and Simoneit2009), but also occurred in other environments, such as nearshore; therefore, fossil charcoal abundance should be taken with some caution as an ‘in situ’ indicator of wildfire frequency (Pieńkowski & Waksmundzka, Reference Pieńkowski and Waksmundzka2009).

In addition, we would like to point out that release of toxic pollutants such as SO₂, sulphate aerosols and PAHs certainly led to defoliation, which increased forest flammability and resulting fire activity, similar to the climate-driven shift from broad-leaved to narrow-leaved taxa at the T–J boundary (McElwain, Wagner & Hesselbo, Reference McElwain, Wagner and Hesselbo2009; Belcher et al., Reference Belcher, Mander, Rein, Jervis, Haworth, Hesselbo, Glasspool and McElwain2010a).

5.b. Carbon isotopes

Terrestrial organic matter displays significant interspecific variation in isotopic composition owing to variations in the organic composition and/or pedogenic processes; thus it may be argued that only a compound-specific analysis is likely to reveal an atmospheric change in isotopic composition (Kürschner, Bonis & Krystyn, Reference Kürschner, Bonis and Krystyn2007; Ruhl, Kürschner & Krystyn, Reference Ruhl, Kürschner and Krystyn2009; Ruhl, Veld & Kürschner, Reference Ruhl, Veld and Kürschner2010). The obtained results reflect carbon-cycle disturbances in the atmospheric system from a limited set of plant organs. Therefore, the isotope data obtained from the palynomaceral separates (i.e. separated phytoclasts) may be the most reliable, although obtained by a very arduous method.

The carbon isotope excursions (CIEs) at the T–J boundary interval are known from many (mostly marine) successions from several continents from low to mid palaeolatitudes, and provide a robust and reproducible means of correlation. The CIE is well documented at several other former GSSP candidates including St Audrie's Bay, UK (Hesselbo et al. Reference Hesselbo, Robinson, Surlyk and Piasecki2002), Larne, Northern Ireland (Simms & Jeram, Reference Simms and Jeram2006), Queen Charlotte Islands (Kunga Island and Kennecott Point combined), British Columbia (Ward et al. Reference Ward, Haggart, Carter, Wilbur, Tipper and Evans2001, Reference Ward, Garrison, Haggart, Kring and Beattie2004; Williford et al. Reference Williford, Ward, Garrison and Buick2007), and numerous sites in Austria, including Kendelbachgraben and Tiefengraben and particularly at Kuhjoch, the GSSP for the base of the Jurassic System (Kürschner, Bonis & Krystyn, Reference Kürschner, Bonis and Krystyn2007; Ruhl et al. Reference Ruhl, Kürschner, Reichart and Krystyn2007; Hillebrandt et al. Reference Hillebrandt, Krystyn, Kürschner, Bown, McRoberts, Uhl, Simms, Tomasovych and Ulrichs2007). At most localities the CIE is divided into two excursions: the lower (‘initial‘) and the upper (‘main’, although not so well characterized), separated by a positive excursion. Of note is also another negative CIE observed in older Rhaetian strata (i.e. the Westbury Fm in St Audrie's Bay, see Hesselbo et al. Reference Hesselbo, Robinson, Surlyk and Piasecki2002). Possibly, the most marked excursion from the Westbury Fm corresponds to the oldest excursion observed in Kamień Pomorski at the depth of 699.3 m (Fig. 6). Furthermore, the position of the sequence boundaries (emergence surfaces or lowstand deposits) in St Audrie's Bay, Kuhjoch and Csövár and the most prominent sequence boundary in Kamień Pomorski at 698 m provide additional correlative constraints, separating the older Rhaetian excursions from subsequent ‘initial’ excursion (Fig. 6). The ‘initial’ excursion, was regarded as so significant, that it prompted the suggestion that isotope stratigraphy may be used to define the position of the T–J boundary (Hesselbo et al. Reference Hesselbo, Robinson, Surlyk and Piasecki2002; Hesselbo, Robinson & Surlyk, Reference Hesselbo, Robinson and Surlyk2004). The advantage of choosing the CIE event is that it has been widely recognized in North America and Europe. However, the isotope excursions have not so far been described from South America. The disadvantage of adopting a carbon isotope marker is that it could become unrecognizable or distorted by the occurrence of minor gaps in the succession or by post-depositional alteration; errors in identification and correlation could occur, so independent biochronological corroboration will in all cases be required. In the UK at St Audrie's Bay, the ‘initial’ excursion is located slightly below the last occurrence of conodonts (top of the Cotham Member of the Lilstock Formation); in eastern North America the ‘initial’ excursion spans most of the prominent evolutionary events in the latest Rhaetian, such as the Newark (Passaic) palynofloral turnover and the first CAMP eruptions (Lucas & Tanner, Reference Lucas and Tanner2007; Korte et al. Reference Korte, Hesselbo, Jenkyns, Rickaby and Spötl2009; Deenen et al. Reference Deenen, Ruhl, Bonis, Krijgsman, Kürschner, Reitsma and Van Bergen2010; Whiteside et al. Reference Whiteside, Olsen, Eglinton, Brookfield and Sambrotto2010). The above observations show that the ‘initial’ carbon isotope excursion is of great correlative significance and it coincides with the onset of magmatism in the CAMP and onset of the Late Rhaetian extinction level (Deenen et al. Reference Deenen, Ruhl, Bonis, Krijgsman, Kürschner, Reitsma and Van Bergen2010; Whiteside et al. Reference Whiteside, Olsen, Eglinton, Brookfield and Sambrotto2010).

In the Kamień Pomorski continental succession, this marked ‘initial’ negative CIE occurs at a depth of ~ 685–692 m, spanning some 7 m of sediments (Fig. 6) and is located within the range of the typical Rhaetian miospores (Fig. 4). Two negative sub-peaks of the ‘initial’ negative excursion, separated by a more positive value, are observed (although only in three samples). Judging from the values of δ¹³C and its stratigraphical position, the ‘bi-partite’ character of the ‘initial’ excursion can be tentatively compared to the isotope curves obtained from many Late Rhaetian marine sections: Pálfy & Dosztály (Reference Pálfy and Dosztaly2000), Pálfy et al. (Reference Pálfy, Demeny, Haas, Hetenyi, Orchard and Veto2001, Reference Pálfy, Demeny, Haas, Carter, Gorog, Halasz, Oravecz-Scheffer, Hetenyi, Marton, Orchard, Ozsvart, Veti and Zajzon2007), Götz et al. (Reference Götz, Ruckwied, Pálfy and Haas2009) in Hungary; Ward et al. (Reference Ward, Garrison, Williford, Kring, Goodwin, Beattie and McRoberts2007) and Hesselbo, McRoberts & Pálfy (Reference Hesselbo, McRoberts and Pálfy2007) in Nevada, USA; Hesselbo et al. (Reference Hesselbo, Robinson, Surlyk and Piasecki2002) in Greenland, and Ruhl, Veld & Kürschner (Reference Ruhl, Veld and Kürschner2010) in the Northern Calcareous Alps (even if the section including the ‘initial’ negative excursion is highly condensed there).

Next, the higher section between 681 and 685 m is characterized by more positive values (−26.0, −26.5 and particularly a −24.0 positive peak at 681.5 m). Above 681.5 m there is a trend back towards more negative values (−25.5, −26.0), with a slight return to more positive values at 676.9–677.5 m. The most positive peak at 681.5 m we correlate with the lower part of the positive excursion at St Audrie's Bay, separating the ‘initial’ and ‘main’ negative isotope excursions (Hesselbo et al. Reference Hesselbo, Robinson, Surlyk and Piasecki2002).

Recently, Korte et al. (Reference Korte, Hesselbo, Jenkyns, Rickaby and Spötl2009, fig. 1c) have correlated the carbon isotope curves from St Audrie's Bay and Kujoch, placing the T–J boundary at a secondary negative peak within the positive isotope excursion, i.e. at the top of Langport Member and Lilstock Formation (Hesselbo et al. Reference Hesselbo, Robinson, Surlyk and Piasecki2002), although Ruhl, Kürschner & Krystyn (Reference Ruhl, Kürschner and Krystyn2009) and Deenen et al. (Reference Deenen, Ruhl, Bonis, Krijgsman, Kürschner, Reitsma and Van Bergen2010) still correlate the T–J boundary at Kujoch with the higher part of the section at St Audrie's (onset of the ‘main’ isotope excursion at St Audrie's or even the higher part of this profile). Based on miospores and megaspore evidence and succession of carbon isotope disturbances, we follow the correlation of Korte et al. (Reference Korte, Hesselbo, Jenkyns, Rickaby and Spötl2009). Consequently, the slight trend towards more negative values above the positive excursion at 681.5 m in Kamień Pomorski (Fig. 7) is correlated with the subordinate negative peak within the positive excursion at St Audrie's, adopted also as the geochemical marker of the T–J boundary (Korte et al. Reference Korte, Hesselbo, Jenkyns, Rickaby and Spötl2009). Again, it should be borne in mind that some section in Kamień Pomorski could be missing at the sequence boundary.

The second negative excursion is probably hidden within the uncored section between 653.5 and 676.5 m and correlation with Hettangian strata in the marine profiles of St Audrie's, New York Canyon and Tyrol is based on the appearance of Cerebropollenites thiergartii and a Jurassic palynomorph assemblage just above the Rhaetic palynomorph assemblage. The upper range of the ‘main’ isotopic excursion is unclear. In England (Hesselbo et al. Reference Hesselbo, Robinson, Surlyk and Piasecki2002), the top of the main excursion has not been documented; in the USA and Tyrol (Ward et al. Reference Ward, Garrison, Williford, Kring, Goodwin, Beattie and McRoberts2007; Hillebrandt et al. Reference Hillebrandt, Krystyn, Kürschner, Bown, McRoberts, Uhl, Simms, Tomasovych and Ulrichs2007; Kürschner et al. Reference Kürschner, Bonis and Krystyn2007; Ruhl et al. Reference Ruhl, Kürschner, Reichart and Krystyn2007) this excursion is terminated within the Lower Hettangian (Psiloceras planorbis chron), while in other profiles from Canada (Williford et al. Reference Williford, Ward, Garrison and Buick2007), this excursion extends stratigraphically higher than any other extant dataset and demonstrates that the ‘main’ isotope excursion does not represent a short event at all, but in fact a prolonged trend of isotopically light organic matter that extends well into the Sinemurian.

5.c. Rhenium/osmium isotopic system and iridium content

Obtained results (Table 1; Fig. 8), showing the decreasing trend of ¹⁸⁷Os/¹⁸⁸Os and ¹⁸⁷Os/¹⁸⁶Os values up to the sequence boundary at 678.4 m and increase (although not so stable) in ¹⁹²Os values indicates an increase in the relative supply rate of non-radiogenic Os from the mantle, most probably associated with emplacement of the CAMP. The subsequent shift towards radiogenic values in Hettangian strata reflects an increased supply of radiogenic Os owing to enhanced continental weathering (Cohen & Coe, Reference Cohen and Coe2002, Reference Cohen and Coe2007; Kuroda et al. Reference Kuroda, Hori, Suzuki, Gröcke and Ohkouchi2010). A marked increase in the most common Os stable isotope (¹⁹²Os), coupled with a decrease in both ¹⁸⁷Os/¹⁸⁸Os and ¹⁸⁷Os/¹⁸⁶Os and δ¹³C values (‘initial’ negative excursion) point to particularly intense volcanic activity at that level. ¹⁸⁷Os/¹⁸⁸Os and ¹⁸⁷Os/¹⁸⁶Os values further decrease to the top of Rhaetian strata (678.6 m), but the increase of ¹⁹²Os is this time slighter and occurs about 1 m higher at 677.5 m. Cohen & Coe (Reference Cohen and Coe2002, Reference Cohen and Coe2007) and Kuroda et al. (Reference Kuroda, Hori, Suzuki, Gröcke and Ohkouchi2010) analysed the Os-isotope system in marine mudrock samples from St Audrie's Bay, UK, and pelagic deposits at Kurusu, Japan. These authors confirmed coincidence between the carbon-cycle disturbances and changes in the Os system reflecting volcanic activity versus continental weathering. Results obtained herein show that similar changes in the Os-isotope system occur also in the continental realm (fluvio-lacustrine environment) and are also linked with the carbon-cycle disturbances (particularly to the ‘initial’ one). In a continental environment, the herein reported decrease in the ¹⁸⁷Os/¹⁸⁸Os and increase in stable ¹⁹²Os (Fig. 8) might point to the volcanic fallout, which must have been extensive enough to reach the Pomerania area, i.e. a distance of some 1500 km (at that time) from the CAMP (Cirilli et al. Reference Cirilli, Marzoli, Tanner, Bertrand, Buratti, Jourdan, Bellieni, Kontak and Renne2009; Muttoni et al. Reference Muttoni, Kent, Jadoul, Olsen, Rigo, Galli and Nicora2010; van de Schootbrugge et al. Reference Van de Schootbrugge, Quan, Lindström, Püttmann, Heunisch, Pross, Fiebig, Petschick, Röhling, Richoz, Rosenthal and Falkowski2009; Deenen et al. Reference Deenen, Ruhl, Bonis, Krijgsman, Kürschner, Reitsma and Van Bergen2010). Based on existing knowledge of the osmium isotopic system in the geological cycle, it is hard to point to an alternative explanation.

Moreover, van de Schootbrugge et al. (Reference Van de Schootbrugge, Quan, Lindström, Püttmann, Heunisch, Pross, Fiebig, Petschick, Röhling, Richoz, Rosenthal and Falkowski2009) indicated that the T–J boundary is marked by an enrichment of PAHs, which, in the absence of correlative charcoal peaks, the authors interpreted as an indication of incomplete combustion of organic matter by ascending flood basalt lava. Van de Schootbrugge et al. (Reference Van de Schootbrugge, Quan, Lindström, Püttmann, Heunisch, Pross, Fiebig, Petschick, Röhling, Richoz, Rosenthal and Falkowski2009) also observed that enrichment of PAHs was associated with darkening of miospores, likely caused by acid rains. The same changes (enrichment in PAHs and darkening of miospores) can be also observed in the Kamień Pomorski profile (Fig. 8). Marynowski & Simoneit (Reference Marynowski and Simoneit2009) performed organic geochemistry analyses in the same samples from Kamień Pomorski. According to them, three levels (one at 687 m within the ‘initial’ δ¹³C excursion, a second interval 676.9–679.4 m from the T–J transitional section and a third Hettangian interval 650–653 m) show elevated PAH concentrations, with phenanthrene, fluoranthene, pyrene and benzo[ghi]perylene being most abundant fractions (Marynowski & Simoneit Reference Marynowski and Simoneit2009, fig. 9 and supplementary data). Although charcoal is present throughout the Kamień Pomorski profile as dispersed, crushed and angular fragments, there are no peculiar charcoal concentrations which might be statistically correlated to the observed PAH peaks.

At the depths 686 m and 678.6 m, darkened (brown and orange-brown) miospores were observed (Figs 3, 8). Collectively, all these indicators (carbon and osmium isotopic system disturbances, darkened miospores and PAH concentrations) tend to support van de Schootbrugge et al. (Reference Van de Schootbrugge, Quan, Lindström, Püttmann, Heunisch, Pross, Fiebig, Petschick, Röhling, Richoz, Rosenthal and Falkowski2009), suggesting that the release of volcanic-derived pollutants such as sulphur dioxide and toxic compounds such as PAHs may have contributed to the observed δ¹³C, PAH and Os geochemical disturbances and resulting extinction. Wildfires would accomplish the whole scenario, although in our opinion PAHs were produced not only by local wildfires. Belcher et al. (Reference Belcher, Mander, Rein, Jervis, Haworth, Hesselbo, Glasspool and McElwain2010a) indicated increased fire activity at the T–J boundary in Greenland due to climate-driven floral changes, which is there additionally supported by a fivefold increase in the abundance of fossil charcoal in the earliest Jurassic. Similarly, wildfires also occurred during the T–J transitional time in Poland, as documented by Marynowski & Simoneit (Reference Marynowski and Simoneit2009) and Pieńkowski & Waksmundzka (Reference Pieńkowski and Waksmundzka2009).

Additionally, the Os isotopic system studied in nine samples (Table 1; Fig. 8) can also be indicative of the potential presence of extraterrestrial matter. The most common criteria applied in such studies are the detection of a positive siderophile element anomaly (enrichment of ejecta fallout matter in Os, Re, Ir and other siderophile elements at the ppb level) and the Re–Os isotopic system (¹⁸⁷Os/¹⁸⁶Os as well as ¹⁸⁷Os/¹⁸⁸Os ratios). As the Earth's crust is characterized by a highly radiogenic ratio of Os isotopes (high ¹⁸⁷Os content) and meteorites are characterized by the opposite (non-radiogenic ratio of Os isotopes), the Os isotopic system is regarded as the most sensitive, thus the most reliable indicator of the presence of extraterrestrial matter in sediments (Geyh & Schleicher, Reference Geyh, Schleicher, Geyh and Schleicher1990; Dickin, Reference Dickin and Dickin1995; Faure, Reference Faure1986; Muñoz-Espadas, Martinez-Frias & Rosario, Reference Muñoz-Espadas, Martinez-Frias, Rosario, Koeberl and Muñoz-Espadas2003). An ¹⁸⁷Os/¹⁸⁶Os ratio of about 1.05–1.1 (Geyh & Schleicher, Reference Geyh, Schleicher, Geyh and Schleicher1990; Dickin, Reference Dickin and Dickin1995) and an ¹⁸⁷Os to ¹⁸⁸Os ratio of about 0.12–0.13 (Faure, Reference Faure1986; Muñoz-Espadas, Martinez-Frias & Rosario, Reference Muñoz-Espadas, Martinez-Frias, Rosario, Koeberl and Muñoz-Espadas2003) point to the presence of extraterrestrial matter (or alternatively, also of mantle-derived material); higher values are regarded as typical of the Earth's crust. This was first shown for the Cretaceous–Palaeogene boundary by Turekian (Reference Turekian1982), Luck & Turekian (Reference Luck and Turekian1983) and later in other analyses (Koeberl & Shirey, Reference Koeberl and Shirey1997; Koeberl, Reference Koeberl, Grady, Hutchison, McCall and Rothery1998). Based on elevated iridium content at the T–J boundary in the Newark Basin, Olsen et al. (Reference Olsen, Kent, Sues, Koeberl, Huber, Montanari, Rainforth, Fowell, Szajna and Hartline2002) hypothesized, that the T–J boundary supposed mass extinction was connected with an impact event and ejecta layer (recently, this view has been softened by Whiteside et al. Reference Whiteside, Olsen, Eglinton, Brookfield and Sambrotto2010).

In the samples studied, all the ¹⁸⁷Os/¹⁸⁶Os values (initial) are between 3.838 and 4.873, and ¹⁸⁷Os/¹⁸⁸Os ratio (initial) ranges between 0.427 and 0.586 (Table 1; Fig. 8). These results indicate that the Os isotopic system of the levels sampled is characteristic of typical crustal rocks, namely for the sediments derived most likely from the weathered Precambrian crystalline rocks. Furthermore, Ir content is less than 5 ppt (measurement sensitivity threshold), except for one sample (depth 703.2 m) where the Ir content was 7 ppt. In conclusion, the samples studied (in respect of the ¹⁸⁷Os/¹⁸⁶Os and ¹⁸⁷Os/¹⁸⁸Os ratio values as well as the Ir content) do not contain detectable extraterrestrial (or larger amounts of mantle) matter, which lends no support to the hypothesis of a large projectile impact at the T–J boundary.