2. Geological setting

The material studied comes from the Albian–Cenomanian interval of the mid-Cretaceous (uppermost lower Albian – lower Turonian) condensed, phosphorite-bearing succession exposed along the limbs of the Annopol anticline, central Poland (Fig. 1a–c). The Annopol anticline is located on the east bank of the Wisła (Vistula) River, central Poland (Fig. 1b). This is the easternmost part of the so-called Mesozoic Border of the Holy Cross Mountains (Marcinowski & Radwański, Reference Marcinowski and Radwański1983).

The lithology of the mid-Cretaceous succession at Annopol was described by Samsonowicz (Reference Samsonowicz1925, Reference Samsonowicz1934), Pożaryski (Reference Pożaryski1947), Cieśliński (Reference Cieśliński1959) and Walaszczyk (Reference Walaszczyk1987). The biostratigraphy was based mainly on ammonites and inoceramid bivalves (Samsonowicz, Reference Samsonowicz1925, Reference Samsonowicz1934; Pożaryski, Reference Pożaryski1947; Cieśliński, Reference Cieśliński1959, Reference Cieśliński1987; Marcinowski, Reference Marcinowski1980; Marcinowski & Radwański, Reference Marcinowski and Radwański1983, Reference Marcinowski and Radwański1989; Marcinowski & Walaszczyk, Reference Marcinowski and Walaszczyk1985; Marcinowski & Wiedmann, Reference Marcinowski and Wiedmann1985, Reference Marcinowski and Wiedmann1990; Machalski & Kennedy, Reference Machalski and Kennedy2013). The reader is referred to Table 1 for the standard ammonite and foraminiferal subdivisions of the Albian–Cenomanian interval which we have used.

Table 1.

Ammonite and foraminiferal zonation of the Albian–Cenomanian interval studied. Ammonite biostratigraphy follows Machalski & Kennedy (Reference Machalski and Kennedy2013) for the Albian stage and Kennedy & Gale (Reference Kennedy and Gale2006) for the Cenomanian stage. Foraminiferal zonation based on Robaszynski & Caron (Reference Robaszynski and Caron1995), Kennedy et al. (Reference Kennedy, Gale, Lees and Caron2004) and Ogg & Hinnov (Reference Ogg, Hinnov, Gradstein, Ogg, Schmitz and Ogg2012).

The Annopol succession is extremely thin when compared to coeval strata in adjacent areas due to its location on a submarine high (Cieśliński, Reference Cieśliński1976; Marcinowski & Radwański, Reference Marcinowski and Radwański1983). Based on the model presented by Juignet (Reference Juignet1980) and Hancock (Reference Hancock1990, Reference Hancock2004), the Annopol succession can be subdivided into a series of thin, transgressive units (Machalski & Kennedy, Reference Machalski and Kennedy2013). These units are capped by layers of reworked phosphatic nodules and clasts, burrowed hardgrounds and omission and/or erosional surfaces (Fig. 1a), which are interpreted as regression maxima (compare Hancock, Reference Hancock1990, fig. 2)

3. Sampled interval

The rock interval sampled for this study comprises six Albian–Cenomanian units, c. 5.5 m in total thickness (Fig. 1a). The lower part of unit 1 (unfossiliferous sands on top of the Jurassic substrate) was accessible and sampled only in the surface locality Kopiec, north of Annopol (Fig. 1b). The remainder of the samples are from the underground phosphate mine Jan 1 at Annopol (Fig. 1b,c). All sampled units are listed below in ascending order, with comments on their lithology, ichnology and macro-fossil biostratigraphy.

Unit 1. Quartz sands with a quartzose sandstone layer in the middle, with a burrowed surface at the top. The total thickness of this unit measured at Kopiec is 2.50 m. The sandstone contains ammonites of the upper lower Albian Otohoplites auritiformis Zone, Hoplites (Hoplites) eodentatus Subzone; see Marcinowski & Wiedmann (Reference Marcinowski and Wiedmann1985, Reference Marcinowski and Wiedmann1990) and discussion in Machalski & Kennedy (Reference Machalski and Kennedy2013).

Unit 2. Quartz sands with glauconite and rare ovoid sandstone nodules near the top, c. 70 cm thick. Unit 2 is capped by a distinct phosphate horizon, including characteristic spindle-shaped phosphatic nodules. In the upper part of unit 2, burrows filled with a marly deposit identical to that of unit 4 are visible (Fig. 2d). The phosphatic horizon contains a biostratigraphically mixed ammonite assemblage, ranging from the middle Albian Hoplites dentatus Zone to the lowermost upper upper Albian Mortoniceras fallax Zone (Marcinowski & Radwański, Reference Marcinowski and Radwański1983, Reference Marcinowski and Radwański1989; Machalski & Kennedy, Reference Machalski and Kennedy2013; Kennedy & Machalski, Reference Kennedy and Machalski2015; see Table 1 for the ammonite zonations used in the present paper).

Figure 2.

Close-up views of the upper Albian interval of the Annopol successions to show the bioturbated nature of the phosphatic layer at the top of unit 3 (a, b) and discrete burrows filled with the Cenomanian marls in this unit (c) and near the top of the underlying unit 2 (d). Underground mine at Annopol (a, b, d) and surface locality Kopiec (c).

Unit 3. Quartz sands with glauconite passing upwards into marly sands with glauconite and phosphates, forming a distinct horizon at the top of the unit; the total thickness of unit 3 is c. 50 cm. In many places, the phosphatic horizon has a highly bioturbated fabric with more or less discrete burrows (Fig. 2a–c). Phosphates contain the upper upper Albian ammonites, most probably of the Mortoniceras perinflatum Zone, preserved as attachment scars on oyster shells (Machalski & Kennedy, Reference Machalski and Kennedy2013) and phosphatic ‘pseudosteinkerns’ (see Machalski & Olszewska-Nejbert, Reference Machalski and Olszewska-Nejbert2016). There is a stratigraphical gap at the top of unit 3, embracing the Praeschloenbachia briacensis Zone (Machalski & Kennedy, Reference Machalski and Kennedy2013; Table 1). Unit 3 and the top of unit 2 were collectively referred to as ‘the Phosphorite Bed’ by previous authors (e.g. Marcinowski & Walaszczyk, Reference Marcinowski and Walaszczyk1985; Marcinowski & Wiedmann, Reference Marcinowski and Wiedmann1985; Walaszczyk, Reference Walaszczyk1987; Marcinowski & Radwański, Reference Marcinowski and Radwański1983, Reference Marcinowski and Radwański1989).

Unit 4. Quartzose marls with glauconite and phosphates in the lower part, c. 170 cm thick. There is a concentration of phosphates and fossils, including belemnite Neohibolites ultimus and bivalve Aucellina in the lower part of the unit, which becomes less detrital and more calcareous towards the top. Several burrows with distinctive white marly infillings with no quartz and glauconite (Fig. 3a–d) were recognized during our study in the lower part of the unit. There is no continuous layer of such a lithology in the succession. We therefore interpret these burrows as remnants of a sedimentary unit originally deposited in the Annopol area, but subsequently removed by erosion. Ammonite assemblage from unit 4 comprises lower Cenomanian taxa (Mantelliceras mantelli and M. dixoni zones, according to Marcinowski & Walaszczyk, Reference Marcinowski and Walaszczyk1985, although no unequivocal dixoni fossils are listed by these authors). No macrofossils were found in the above-mentioned burrows.

Figure 3.

(a–d) Lower upper Cenomanian burrows in the lower part of the lower Cenomanian unit 4. Underground mine at Annopol.

Unit 5. Glauconitic marls filling large burrows in the underlying unit, truncated by a complex hardground with strong phosphate and glauconite impregnation (Fig. 1a). This is a highly fossiliferous level, yielding inter alia numerous phosphatized moulds of an irregular echinoid Holaster subglobosus (Samsonowicz, Reference Samsonowicz1925, Reference Samsonowicz1934). According to Marcinowski & Walaszczyk (Reference Marcinowski and Walaszczyk1985) the ammonite assemblage from unit 5, represented exclusively by phosphatized specimens, comprises both lower and middle Cenomanian taxa, the latter characteristic of the Acanthoceras rhotomagense Zone and Turrilites costatus and T. acutus subzones (see Table 1). A stratigraphic gap embracing the A. jukesbrownei Zone (see Table 1) was postulated for the top of unit 5 by Marcinowski (Reference Marcinowski1980). There is also no indication of the Cunningtoniceras inerme Zone (Table 1) fossils, suggesting a similarly significant stratigraphic gap near the bottom of unit 5.

Unit 6. Strongly glauconitic marls with burrowed omission surface at the top, variable in thickness over the range 0–40 cm. The ammonite species Schloenbachia lymensis was quoted by Marcinowski & Walaszczyk (Reference Marcinowski and Walaszczyk1985) from this unit as an upper Cenomanian biostratigraphic marker. According to Kennedy (Reference Kennedy2013, p. 466), this species ‘first appears in the Middle Cenomanian Acanthoceras jukesbrownei Zone, and has its acme in the lower upper Cenomanian Calycoceras guerangeri Zone’. Inoceramids and belemnites reported by Marcinowski & Walaszczyk (Reference Marcinowski and Walaszczyk1985) locate unit 6 in the upper Cenomanian Substage.

The Turonian interval, not sampled here, consists of limestone bed with rare phosphatic nodules and glauconite, capped by a burrowed hardground with glauconitic mineralization (unit 7), followed by marls with a burrowed omission surface at their top (unit 8). These are overlain by marls with layers of black flint (unit 9, visible in the ceiling of the underground corridor in Fig. 1c). Units 7 and 8 represent the lower Turonian Inoceramus labiatus Zone whereas unit 9 belongs to the middle Turonian I. lamarcki Zone (Marcinowski & Walaszczyk, Reference Marcinowski and Walaszczyk1985).

5. Benthic foraminifera

Selected examples of the benthic foraminifera from the Annopol succession are illustrated in Figure 4. The assemblages from the upper part of unit 1 (samples 1–4 in Fig. 5) comprise only agglutinated forms, mainly Arenobulimina and Ataxophragminum. The relative abundance of specific groups of foraminiera recorded in these assemblages is not presented in Figure 5 since foraminifera are infrequent and the number of specimens is less than 100 per 200 g sample.

Figure 4.

Benthic foraminifera from the Annopol succession. (a1, a2) Lingulogavelinella orbiculata (Kusnezova), sample 15, MWGUW ZI/67/01.01. (b1–b3) Lingulogavelinella spinosa (Plotnikova), sample 15, MWGUW ZI/67/01.12. (c1–c3) Lingulogavelinella globosa (Brotzen), sample 23, MWGUW ZI/67/03.17. (d1, d2) Lingulogavelinella kaptarenkae (Plotnikova), sample 13, MWGUW ZI/67/01.05. (e1–e3) Orithostella formosa (Brotzen), sample 14, MWGUW ZI/67/01.07. (f1–f3) Berthelina intermedia (Berthelin), sample 17, MWGUW ZI/67/01.08. (g1–g3) Gavelinella berthelini (Keller), sample 23, MWGUW ZI/67/03.21. (h1–h3) Berthelina baltica (Brotzen), sample 16, MWGUW ZI/67/01.44. (i1–i3) Berthelina cenomanica (Brotzen), sample 20, MWGUW ZI/67/01.03. Scale bars = 100 µm.

Figure 5.

Changes in benthic foraminiferal assemblages in the Albian–Cenomanian interval in the Annopol succession. A – proportions of calcareous and agglutinated forms within benthic foraminiferal assemblages; B – relative abundances of dominant and common species or groups of species; 1–23 – sample position.

The assemblages from samples 5–7 taken from unit 2 (Fig. 5) are also characterized by abundant agglutinated foraminifera; however, calcareous benthic foraminifera appear as an increasingly important element. Calcareous foraminifera are dominated by relatively small and thin-walled representatives of Orithostella (O. formosa) and Lingulogavelinella (the most numerous are L. kaptarenkae, L. spinosa, L. orbiculata and L. varsoviensis). The proportional abundance of calcareous foraminifera in relation to agglutinated forms successively increases upwards in the unit.

Samples 10, 11 and 13–15 from unit 3 (Fig. 5) yield highly diverse assemblages represented by more than 30 species. The most abundant forms belong to the following genera: Berthelina (B. baltica, B. belorussica, B. intermedia, B. cenomanica, B. gorzowiensis); Cibicides (C. gorbenkoi); Gavelinella (G. schloenbachi); Gyroidinoides (G. infracretacea); Lenticulina (L. rotulata, L. muensteri, L. spp.); Lingulogavelinella (L. kaptarenkae, L. orbiculata, L. spinosa, L. varsoviensis); Orithostella (O. formosa); Praebulimina (P. evexa, P. reussi); Tappanina (T. eouvigeriniformis); and Arenobulimina (A. advena, A. chapmani, A. conoidea, A. frankei, A. obliqua, A. polonica, A. truncata, A. varsoviensis). There is no domination of any specific taxon in this assemblage. Very similar foraminiferal assemblages occur in the uppermost part of unit 2 (sample 9), in the lower part of unit 4 (samples 16, 17 in Fig. 5) and in burrows descending from unit 4 to unit 3 (sample 12 in Fig. 5) and to the upper part of unit 2 (sample 8 in Fig. 5). The vertical range of these assemblages, which seems to be expanded by burrowing organisms, therefore broadly corresponds to that of the phosphates (Fig. 5).

Benthic foraminifera from samples 18, 20, 21 in unit 4 (Fig. 5) are still dominated by calcareous forms; however, in contrast to the assemblages of unit 3, there is clear domination of a single genus Berthelina. This genus encompasses about 40% of the total assemblage and is represented by the following species: B. baltica, B. intermedia, B. cenomanica, B. gorzowiensis and B. belorussica.

Sample 22 from unit 5 (Fig. 5) strongly resembles those from unit 4 in terms of composition, with the exception that Orithostella, a very common taxon in the lower part of the Annopol succession, is entirely absent here. Benthic assemblages from the marly infillings of burrows in unit 4 (sample 19 in Fig. 5) are very similar in composition.

The assemblage from sample 23 of unit 6 (Fig. 5) differs significantly from all the above-mentioned assemblages by the disappearance of many common species (e.g. Berthelinia baltica, B. belorussica, B. intermedia, B. cenomanica, B. gorzowiensis, Lingulogavelinella kaptarenkae, L. orbiculata, L. spinosa and L. varsoviensis). It is also characterized by a predominance of a few, newly appearing species such as Lingulogavelinella globosa and Gavelinella berthelini, which are accompanied by a smaller proportion of other genera, mainly Lenticulina, Gyroidinoides and Valvulinera.

6. Planktonic foraminifera

Selected examples of planktonic foraminifera from the Annopol succession are illustrated in Figures 6 and 7. These foraminifera are entirely absent from unit 1 (samples 1–4 in Fig. 8). In the lower part of unit 2 (samples 5–7 in Fig. 8), they are more frequent than the benthic foraminifera; the relative proportion of planktonic and benthic foraminifera (P/B ratio) amounts to c. 90%. These assemblages are dominated by small and simple forms belonging to two morphogroups: triserial (Guembelitria cenomana) and low-trochospiral with globular chambers (Hedbergella delrioensis, H. planispira, H. infracretacea and Globigerina graysonensis). Planispiral forms (Globigerinelloides bentonensis) and serial heterohelicids (Heterohelix moremani) form only a few percent of the assemblage, whereas keeled and high-trochospiral forms are absent.

Figure 6.

Planktonic foraminifera from the Annopol succession. (a1, a2) Thalmanninella appenninica (Renz), sample 16, MWGUW ZI/67/01.48. (b1–b3) Thalmanninella tehamensis (Marianos and Zingula), sample 13, MWGUW ZI/67/01.50. (c1, c2) Thalmanninella gandolfi (Luterbacher and Premoli-Silva), sample 13, MWGUW ZI/67/03.03. (d1–d3) Thalmanninella globotruncanoides (Sigal), sample 16, MWGUW ZI/67/01.08. (e1–e3) Rotalipora cushmani (Morrow), sample 23, MWGUW ZI/67/03.09. (f1, f2) Thalmanninella deckeii (Franke), sample 22, MWGUW ZI/67/03.016. (g1, g2) Thalmanninella greenhornensis (Morrow), sample 19, MWGUW ZI/67/03.22. (h1–h3) Praeglobotruncana stephani (Gandolfi), sample 23, MWGUW ZI/67/03.03. (i1, i2) Praeglobotruncana gibba (Klaus), sample 19, MWGUW ZI/67/03.15. (j) Guembelitria cenomana (Keller), sample 6, MWGUW ZI/67/02.06. (k) Heterohelix moremani (Cushman), sample 9, MWGUW ZI/67/01.57. Scale bars = 100 µm.

Figure 7.

Planctonic foraminifera from the Annopol succession. (a1, a2) Hedbergella infracretacea (Glaessner), sample 7, MWGUW ZI/67/02.02. (b1–b3) Hedbergella delrioensis (Carsey), sample 13, MWGUW ZI/67/01.37. (c1, c2) Whiteinella brittonensis (Loeblich & Tappan), sample 23, MWGUW ZI/67/03.19. (d1–d3) Whiteinella archaeocretacea Pessagno, sample 23, MWGUW ZI/67/03.20. (e1, e2) Whiteinella aprica (Loeblich & Tappan), sample 23, MWGUW ZI/67/03.01. (f1–f3) Dicarinella hagni (Scheibnerova), sample 19, MWGUW ZI/67/03.27. (g1, g2) Dicarinella sp., sample 19, MWGUW ZI/67/03.28. (h1–h3) Dicarinella imbricata (Mornod), sample 23, MWGUW ZI/67/03.29. (i1–i3) Dicarinella algeriana (Caron), sample 23, MWGUW ZI/67/03.24. (j1, j2) Globigerinelloides bentonensis (Morrow), sample 9, MWGUW ZI/67/01.36. Scale bars = 100 µm.

Figure 8.

Changes in planktonic foraminiferal assemblages in the Albian–Cenomanian interval in the Annopol succession. A – relative abundances of planktonic and benthic foraminifera within foraminiferal assemblages; B – relative abundances of the seven morphotypes of planktonic foraminifera within planktonic foraminiferal assemblages.

In the uppermost part of unit 2 and in unit 3 (samples 9–15 in Fig. 8) planktonic foramiferal assemblages are still characterized by very high P/B ratio (about 80%). However, they differ from those from below in terms of specific morphotype ratios. Triserial forms (Guembelitria) consist of over a dozen percent of the assemblage and globular hedbergellids comprise c. 80%. The complex keeled forms sporadically occur (sample 13); however, they were probably introduced downwards into the older sediments by the activity of burrowing infauna (see discussion in Section 7).

In contrast to the previous assemblages, planktonic foraminiferal assemblages from the lower part of unit 4 (samples 16–17 in Fig. 8) are characterized by the occurrence of more complex morphogroups (single-keeled morphotype (Thalmanninella)) and high trochospiral with globular or narrowed chambers (Whiteinella and Praelobotruncana), still forming just a few percent (1–3%) of the whole assemblage. Very similar planktonic assemblages are recorded in the burrows descending from unit 4 to to unit 3 (sample 12 in Fig. 8) and even to the top part of unit 2 (sample 8 in Fig. 8).

Planktonic foraminiferal assemblages in the higher part of unit 4 (samples 18, 20, 21 in Fig. 8) display a lower P/B ratio which amounts to c. 20–30%. These assemblages are characterized by the more frequent occurrence of the complex planktonic forms.

Aassemblages of planktonic foraminifera from unit 5 (sample 22 in Fig. 8) are characterized by the appearance of single-keeled Rotalipora cushmani. The presence of this taxon is the main difference between unit 5 and the preceding unit.

Planktonic foraminiferal assemblages of marly infillings of burrows within unit 4 (sample 19) display a very high P/B ratio (about 90%) and more abundant complex forms: single-keeled forms are represented by Thalmanninella greenhornensis and Rotalipora cushmani; high trochospiral forms with a pseudo-keel are represented by Praeglobotruncana delrioensis, P. gibba and P. stephani; and high trochospiral forms with simple globular chambers are represented by Whiteinella baltica and W. brittonensis. The assemblages are also characterized by the first appearance of early double-keeled foraminifera of the genus Dicarinella (D. algeriana, D. hagni and D. imbricata). All these complex forms constitute c. 10% of the material.

Planktonic assemblages from unit 6 (sample 23 in Fig. 8) are characterized by a very high P/B ratio (c. 90%) and the frequent occurrence of single-keeled Rotalipora (represented only by R. cushmani), double-keeled Dicarinella (represented by D. algeriana, D. hagni, D. imbricata D. longoriai) and the huge complex Whiteinella (W. aprica, W. archaeocretacea, W. baltica and W. brittonensis) (Fig. 7). They also differ from other assemblages in that they are dominated by high-trochospiral whiteinellids and biserial heterohelicids.

7. Stratigraphy

This is based mainly on planktonic foraminifera, which are one of the biostratigraphic tools for the Albian and Cenomanian stages (Robaszynski & Caron, Reference Robaszynski and Caron1979, Reference Robaszynski and Caron1995; Caron, Reference Caron, Bolli, Saunders and Nielsen1985; Robaszynski et al. Reference Robaszynski, Caron, Amedro, Dupuis, Hardenbol, Gonzales Donoso, Linares and Gartner1993).

Units 1 and 2 yielded no planktonic foraminifera which could be safely interpreted as in situ forms. Well-preserved planktonic forms occur higher in the section and include index species used for standard planktonic foraminiferal zonation (Robaszynski & Caron, Reference Robaszynski and Caron1995), allowing for the correlation of the studied strata with global reference sections. Ranges of the stratigraphically important foraminifera from the Albian–Cenomanian interval at Annopol are presented in Figure 9.

Figure 9.

Stratigraphical distribution of selected planktonic and benthic foraminifera in the Annopol succession.

Sample 13, taken from the middle part of unit 3, yielded the stratigraphicaly important planktonic foraminifera Thalmanninella tehamensis, T. apenninica, T. globotruncanoides and T. gandolfii. The global Boundary Stratotype Section and Point (GSSP) for the base of the Cenomanian Stage was defined (Gale et al. Reference Gale, Kennedy, Burnett, Caron and Kidd1996; Kennedy et al. Reference Kennedy, Gale, Lees and Caron2004) at Mont Risou, France, at the level of the first appearance of a planktonic foraminifer Thalmanninella globotruncanoids (T. brotzeni is a synonym). This level is placed slightly lower than the lowest occurrence of Mantelliceras mantelli, the basal Cenomanian ammonoid marker (Kennedy et al. Reference Kennedy, Gale, Lees and Caron2004). According to the current definition of the GSSP, the occurrence of T. globotruncanoides in unit 3 would indicate that at least part of this unit already represents the lower Cenomanian Substage (Thalmanninella globotruncanoides Zone; see Table 1). There are, however, some important reasons to believe that whole of unit 3 is still of late Albian age and that the younger foraminifera were introduced here by burrowing animals, as suggested earlier by Walaszczyk (Reference Walaszczyk1987). Firstly, several burrows filled with Cenomanian marl identical to that of unit 4 are present in unit 3 and even in the upper part of unit 2 (sample 8, 12; Fig. 2c, d). Marly infills of these burrows yielded the T. appenninica and T. globotruncanoides and it is possible that these taxa have also been locally introduced by burrowers into the sediment which forms our sample 13, albeit in the latter case burrowers did not leave distinct tunnels. The chaotic and patchy distribution of phosphates does suggest strong bioturbation of that level (Fig. 2a, b). Moreover, there are no Cenomanian ammonites in unit 3; all the ammonites present indicate a late Albian age (Machalski & Kennedy, Reference Machalski and Kennedy2013).

Unit 4 yielded the biostratigraphically important Praeglobotruncana stephani, Thalmanninella appenninica, T. globotruncanoides which confirm the lower Cenomanian position of this unit as indicated by macro-fossils (the Thalmanninella globotruncanoides Zone of Reference PerytPeryt, 1983a , b; see also Walaszczyk, Reference Walaszczyk1987). The early Cenomanian age is supported by the occurrence of benthic foraminifer Orithostella formosa (Orithostella jarzevae is a synonym), regarded as a marker for the lower Cenomanian deposits in western Europe (Hart et al. Reference Hart, Bailey, Crittenden, Fletcher, Price, Świecicki, Jenkins and Murray1989). The last appearance of O. jarzevae in southern England is recorded in the middle of the lower Cenomanian deposits within the Mantelliceras saxbii Subzone (Table 1). The disappearance of this taxon in the lower Cenomanian succession was also noted in western Ukraine (Z. Dubicka, unpubl. PhD thesis, Polish Academy of Sciences, 2012). At Annopol it occurs close to the top of unit 4 (Fig. 9), adding to our doubts about the presence of the M. dixoni Zone in that interval.

Marly fills of burrows in the lower part of unit 4 (sample 19), interpreted as remnants of a lost sedimentary unit, yield the following stratigraphically important planktonic species: Praeglobotruncana gibba, P. stephani, Thalmanninella greenhornensis, Rotalipora cushmani, Whiteinella baltica and W. brittonensis. Early dicarinellids without fully developed keel structures are also present. These are Dicarinella algeriana, D. hagni, D. imbricata and species of the benthic Berthelina intermedia group (B. intermedia, B. baltica and B. cenomanica). All these taxa indicate the lower part of the upper Cenomanian Substage, namely the upper part of R. cushmani Zone. Based on comparison with the foraminiferal assemblages recorded from the southern England (Paul et al. Reference Paul, Lamolda, Mitchell, Vaziri, Gorostidi and Marshall1999), these infillings correspond to an upper Cenomanian interval ranging from the Grey Chalk Formation (Calycoceras guerangeri Zone and the ‘pre-excursion phase of δ13 C’) to the lowermost Plenus Marl (bed 1a) (Metoicoceras geslinianum Zone) of the global reference section for the Cenomanian–Turonian interval in Eastbourne, Sussex, SE England.

Unit 5 is characterized by the co-occurrence of Praeglobotruncana gibba, Rotalipora cushmani and Thalmanninella deeckei indicating the middle Cenomanian unit (Caron, Reference Caron, Bolli, Saunders and Nielsen1985; Robaszynski et al. Reference Robaszynski, Caron, Amedro, Dupuis, Hardenbol, Gonzales Donoso, Linares and Gartner1993), in concert with ammonite dating and the foraminiferal data presented by Marcinowski and Walaszczyk (Reference Marcinowski and Walaszczyk1985) and Walaszczyk (Reference Walaszczyk1987). There is no evidence for the presence of the Thalmanninella reicheli Zone (Table 1) in the section, in agreement with Reference PerytPeryt (1983a , b). As the ammonites of the equivalent Cunningtoniceras inerme Zone (Table 1) are also missing, this suggests another significant gap in the succession.

Unit 6 yields important Dicarinella (D. algeriana, D. hagni and D. imbricata) and Praeglobotruncana (Praeglobotruncana gibba and P. stephani), a single species of one-keeled forms (Rotalipora cushmani) as well as abundant large whiteinellids (Whiteinella baltica, W. brittonensis, W. aprica and W. archaeocretacea). Benthic foraminifera are characterized by the absence of the Berthelina intermedia group and by the occurrence of Lingulogavelinella globosa and Gavelinella berthelini. These taxa indicate the uppermost part of the Rotalipora cushmani Zone (Table 1) and correspond to the interval from bed 2 to bed 4 in the Plenus Marl (Metoicoceras geslinianum ammonite Zone) as exposed in the Eastbourne succession (Paul et al. Reference Paul, Lamolda, Mitchell, Vaziri, Gorostidi and Marshall1999; Pearce, Jarvis & Tocher, Reference Pearce, Jarvis and Tocher2009).

8. Palaeoenvironment

Foraminifera are very useful proxies for palaeoenvironmental studies, based on the correlation between their test shape, composition and environmental requirements (Corliss, Reference Corliss1985; Murray, Reference Murray2006). Almost the entire Annopol succession contains mineral-walled foraminifera, ascribed to the agglutinated and/or calcareous foraminifera. No foraminifera are recorded from sands representing the lower part of unit 1.

Mineral-walled foraminifera are characteristic of marine or brackish environments. Today, the transition from marine to freshwater environments is marked by an almost total disappearance of foraminifers, except for the organic-walled Allogromiida (Sen Gupta, Reference Sen Gupta2002). Accordingly, the barren sands near the base of the Annopol mid-Cretaceous succession could even represent a non-marine environment.

The upper part of unit 1 contains mineral-walled foraminifera, which are represented exclusively by agglutinated taxa. Present-day benthic assemblages from very shallow, sandy-bottom environments are typified by an abundance of agglutinated taxa, which show a preference for sands and silts (Abou Ouf, Reference Abou Ouf1992). For example, the foraminiferal assemblages collected at the depth interval 5–50 m off the sandy coast of the Sea of Japan are dominated by agglutinated Ammobaculites, Eggerellina and Textularia (Matoba, Reference Matoba, Takayanagi and Saito1976). Likewise, the assemblages from the sandy facies off the Washington continental shelf are dominated by agglutinated Eggerellina and Spiroplectammina (Snyder, Hale & Kontrovitz, Reference Snyder, Hale and Kontrovitz1990). In recent nearshore environments, planktonic foraminifera may be entirely absent (Wang, Zhang & Min, Reference Wang, Zhang, Min and Wang1985). Extremely shallow-marine conditions, probably a near-sandy-beach setting, are therefore postulated here for the interval discussed. This is supported by the presence of Ophiomorpha burrows in sandstones in the middle part of unit 1 (Fig. 1a). These burrows, produced by the calianassid shrimps, point to deposition of sands in an extremely shallow-marine, marginal environment (compare Radwański et al. Reference Radwański, Wysocka and Górka2012).

Planktonic as well as calcareous benthic foraminifera appear in the upper part of unit 2, indicating a sea-level rise. However, planktonic foraminifera are represented only by shallow-water dwellers: low-trochospiral hedbergellids, biserial heterohelicids and triserial guembelitrids (see Bé, Reference Bé and Ramsey1977; Caron & Homewood, Reference Caron and Homewood1983; Leckie, Reference Leckie1987). The benthic foraminifera are characterized by a high proportion of agglutinated forms and small, thin-walled, low-diversity calcareous taxa. This still points to a very shallow shelf environment (a few dozen metres deep). A very high P/B ratio is measured for the samples from the upper part of unit 2 (samples 6, 7 and 9), which is usually characteristic of a much deeper shelf environment; however, these results are from a relatively low presence of benthic foraminifera. The successive increase in the relative abundance of calcareous benthic foraminifera through unit 2 suggests a slight, progressive sea-level rise.

In the uppermost part of unit 2 and unit 3 (the Phosphorite Bed of other authors), the most shallow-dwelling planktonic foraminifera (triserial Guembelitria) decrease in abundance in relation to the deeper Hedbergella. A decrease in the abundance of benthic agglutinated forms is also noted. These changes indicate a transgressive pulse. The high diversity of benthic foraminifera suggests substantially higher food availability in comparison to the preceding units. A very high P/B ratio suggests a bloom of planktonic foraminifera, dominated by small, globular hedbergellids. The latter are interpreted as r-strategists, thriving in nutrient-rich waters (Premoli Silva & Sliter, Reference Premoli Silva, Sliter, Barrera and Johnson1999; Petrizzo, Reference Petrizzo2002). Walaszczyk (Reference Walaszczyk1987) interpreted the rise in frequency of hedbergellids as a result of high organic production in the surface waters. Accordingly, foraminiferal assemblages from the Phosphorite Bed, both planktonic and benthic, are interpreted here as the high productivity assemblages sensu Gooday (Reference Gooday2003). These are characteristic of environments with a high and relatively continuous input of organic matter, supplied from intense primary production associated with upwelling, hydrographic fronts or major river discharges (Gooday, Reference Gooday2003). Summarizing, benthic and planktonic foraminifers suggest a high rate of primary production combined with sea-level rise for this part of the succession.

In units 4 and 5, deep-dwelling Thalmanninella, Praeglobotruncana and Whiteinella become common (their sporadic occurrence in the lower levels is ascribed to burrowing activity of macrobenthic organisms; see above). This change is most likely a consequence of a further transgressive pulse. However, these complex foraminifera are still infrequent. Additionally, a relatively low P/B ratio (20–30%) points to a moderate water depth (as for the Cretaceous epicontinental seas). Benthic foraminifera are here dominated by huge, calcareous, epifaunal forms (Berthelina), regarded as typical of oxic environments (Kaiho, Reference Kaiho1994, Reference Kaiho1999). The contribution of agglutinated and infaunal taxa is much smaller. Accordingly, these sediments were probably deposited in mesotrophic and well-oxygenated bottom water conditions with a moderate input of terrigenous material (see Jorissen, Stigter & Widmark, Reference Jorissen, Stigter and Widmark1995).

Marly infillings of burrows in the lower part of unit 4 (sample 19) display a significant increase in the relative abundance of deep-dwelling planktonic foraminifera and a much higher P/B ratio (around 80%) in relation to the older units 4 and 5. These changes probably indicate that a pronounced sea-level rise occurred during the late part of the late Cenomanian Age (upper part of R. cushmani Zone).

Foraminiferal assemblages from unit 6 show a very high P/B ratio with abundant deep-dwelling forms, interpreted by Walaszczyk (Reference Walaszczyk1987) as a result of another, major transgressive pulse. This very high P/B ratio is additionally caused by the radical decrease in number and diversity of benthic foraminifera (Walaszczyk, Reference Walaszczyk1987). Among the latter, the disappearance of the Berthelina intermedia group and appearance of a low-diversity assemblage dominated by Gavelinella berthelini and Lingulogavelinella globosa are observed. This foraminiferal crisis is linked with the Cenomanian–Turonian Boundary Event as recorded in southern England (Jarvis et al. Reference Jarvis, Carson, Cooper, Hart, Leary, Tocher, Horne and Rosenfeld1988; Hart, Dodsworth & Duane, Reference Hart, Dodsworth and Duane1993; Paul et al. Reference Paul, Lamolda, Mitchell, Vaziri, Gorostidi and Marshall1999), SE France (Tronchetti & Grosheny, Reference Tronchetti and Grosheny1991), Poland (Leary & Peryt, Reference Leary and Peryt1991; Peryt & Wyrwicka, Reference Peryt and Wyrwicka1991, Reference Peryt and Wyrwicka1993; Peryt et al. Reference Peryt, Wyrwicka, Orth, Attrep and Quintana1994) and western Ukraine (Z. Dubicka, unpubl. PhD thesis, Polish Academy of Sciences, 2012).

The spreading of the oxygen-minimum bottom waters was commonly invoked as a major cause of extinction of Cenomanian benthic foraminifera (Jarvis et al. Reference Jarvis, Carson, Cooper, Hart, Leary, Tocher, Horne and Rosenfeld1988; Peryt & Wyrwicka, Reference Peryt and Wyrwicka1991; Tronchetti & Grosheny, Reference Tronchetti and Grosheny1991; Hart, Dodsworth & Duane, Reference Hart, Dodsworth and Duane1993; Peryt et al. Reference Peryt, Wyrwicka, Orth, Attrep and Quintana1994). Based on the assumption that the oxygen-minimum zone occurred at the sea bottom during the Oceanic Anoxic Event 2 (OAE2), benthic species and morphotypes which survived this event were considered to be resistant to oxygen-depleted environments and even regarded as their potential indicators. However, Gale et al. (Reference Gale, Smith, Monks, Young, Howard, Wray and Huggett2000) argued that there is no evidence for the development of the oxygen-minimum bottom waters on the European shelf during the Cenomanian–Turonian transition.

According to Gale et al. (Reference Gale, Smith, Monks, Young, Howard, Wray and Huggett2000), a major latest Cenomanian sea-level rise resulted in the spread of oligotrophic, stratified oceanic water onto the shelves. Similarly, Pearce, Jarvis & Tocher (Reference Pearce, Jarvis and Tocher2009) argued that marine productivity collapsed during the latest Cenomanian Age, based on the dinoflagellate evidence. Our data from Annopol support these interpretations. The benthic foraminiferal assemblage from unit 6 is of low abundance and diversity, being dominated by Lingulogavelinella globosa and Gavelinella berthelini. These taxa possess huge (>300 μm), rounded, biconvex, low-trochospiral calcareous tests and are commonly regarded as epifaunal (Corliss & Chen, Reference Corliss and Chen1988; Nagy et al. Reference Nagy, Gradstein, Kaminski and Holbourn1995). According to Kaiho (Reference Kaiho1994, Reference Kaiho1999), Jorissen, Stigter & Widmark (Reference Jorissen, Stigter and Widmark1995), van der Zwaan et al. (Reference Van Der Zwaan, Duijnstee, Den Dulk, Ernst, Jannink and Kouwenhoven1999) and Dubicka & Peryt (Reference Dubicka and Peryt2012), benthic foraminiferal assemblages of low abundance and diversity, dominated by calcareous epifaunal morphotypes, indicate a well-oxygenated oligotrophic environment on the seafloor.

9. Sea-level changes

Walaszczyk (Reference Walaszczyk1987, fig. 4) first published a bathymetric diagram for the Annopol succession, based on the foraminiferal assemblages. His graph reveals a deepening-upwards trend, punctuated by breaks corresponding to the sedimentary discontinuities. This author explained the facies development of the Annopol succession in terms of eustasy rather than local tectonic movements suggested earlier by Cieśliński (Reference Cieśliński1976) and Marcinowski & Radwański (Reference Marcinowski and Radwański1983).

We support the general conclusions of Walaszczyk (Reference Walaszczyk1987) and concur with the eustatic interpretation of the succession (Fig. 10a, b). Proceeding one step further, we attempt here to assess how many of the global mid-Cretaceous sea-level changes can be seen at Annopol. We have chosen a classic graph of the Cretaceous sea-level changes for the British region (Hancock, Reference Hancock1990, fig. 11) for comparison with the bathymetric patterns which can be deduced from the Annopol succession. The relevant interval of the British graph is reproduced here in Figure 10b (note that it preserves original zonation used by that author, which is different in some detail from that used here; see Table 1).

Figure 10.

Interpretation of the Annopol succession in terms of sedimentary cycles, bathymetry and biological productivity. Inserted to the left is a fragment of the graph showing sea-level fluctuations for British region (taken from Hancock, Reference Hancock1990, fig. 11) with our correlation of the patterns seen at Annopol. TUR – Turonian; D – sedimentary discontinuities.

The British curve, thought to reflect a global eustatic pattern, is based on a simple concept of peaks of transgression and troughs of regression (Hancock, Reference Hancock1990, fig. 2; Hancock, Reference Hancock2004). The basic idea behind this concept is that levels of nodular chalks and hardgrounds (in the Chalk) and courses of the phosphatic nodules (in the Gault) define the regressive troughs, whereas the mid-points between these levels reflect transgressive peaks (Hancock, Reference Hancock1990). In practice, the situation is not so simple due to erosion which commonly removed much of the transgressive record during the subsequent regressive phases, particularly in marginal environments. At Annopol for example, only the lowermost parts of the successive units seem to be preserved due to severe erosion during successive sea-level drops. The reader is referred to Ernst et al. (Reference Ernst, Niebuhr, Wiese and Wilmsen1996, fig. 4) for a translation of Hancock's nomenclature into that of the standard sequence stratigraphy and to Haq (Reference Haq2014) for the most recent overview of the Cretaceous sequence stratigraphy and eustasy.

The lowest level at Annopol which can be confidently correlated with the British graph is a discontinuity between units 2 and 3 (marked D2/3 in Fig. 10a). The phosphatic layer at the top of unit 2 contains a mixture of ammonites indicating condensation of several Albian ammonite zones, ranging from Hoplites dentatus Zone to Mortoniceras fallax Zone (Table 1). The latter, being the youngest, dates the phosphatic layer (Kennedy & Machalski, Reference Kennedy and Machalski2015). Nodules yielding these ammonites are probably reworked from several originally separate levels, possibly reflecting several regressive troughs in the middle–lower upper Albian part of the British curve (Fig. 10b). The phosphatic bed at the top of unit 2 corresponds to a regressive trough at the base of the classic Stoliczkaia dispar Zone (Hancock, Reference Hancock1990, fig. 11; Fig. 10b). Reworked phosphatic lags at this position are widespread in Europe (Amédro, Reference Amédro2002). In terms of sequence stratigraphy, the discontinuity between units 2 and 3 at Annopol (D2/3 in Fig. 10a) corresponds to sequence boundary KAl7 of Haq (Reference Haq2014, figs 1, 2). Unit 3 contains late late Albian ammonites, indicative of the M. perinflatum Zone; it is therefore a level within the classic upper upper Albian Stoliczkaia (Stoliczkaiella in current nomenclature) dispar Zone (Machalski & Kennedy, Reference Machalski and Kennedy2013). This is the Vraconnien sensu Amédro (Reference Amédro2002). The foraminifera indicate a transgressive pulse and, indeed, equivalents of this unit occasionally overlap the Jurassic substrate in some localities close to the study area (Pożaryski, Reference Pożaryski1947). On the other hand, phosphates forming the upper part of unit 3 originated in regressive conditions during at least two phases of intraformational reworking or winnowing of fine matrix and concentration of the phosphatic material on the sea bottom (Machalski & Olszewska-Nejbert, Reference Machalski and Olszewska-Nejbert2016). This transgressive–regressive situation is compatible with a stillstand during most of the Stoliczkaia dispar Zone marked on the Hancock graph (Fig. 10b).

Another correlation point is discontinuity D3/4 (Fig. 10a). This is the most prominent stratigraphic boundary in the Annopol succession, separating the Lower and Upper Cretaceous strata. No ammonites indicative of the uppermost Albian – lowermost Cenomanian Praeschloenbachia briacensis Zone are recorded here. The hiatus at the Albian–Cenomanian boundary therefore embraces this zone, a situation analogous to condensed sections elsewhere (see Machalski & Kennedy, Reference Machalski and Kennedy2013). This gap would clearly correspond to the turning point between the stillstand phase for the Stoliczkaia dispar Zone and the beginning of the Cenomanian transgression, marked by Hancock (Fig. 10b). Discontinuity D3/4 at Annopol (Fig. 10a) may be correlated with sequence boundary KAl8 of Haq (Reference Haq2014).

Unit 4 begins the Cenomanian sedimentation and is clearly transgressive, starting with a basal layer of phosphates and fossils (the Neohibolites ultimus/Aucellina Event of Ernst, Schmid & Seibertz, Reference Ernst, Schmid and Seibertz1983; see also Wilmsen, Reference Wilmsen2003, Reference Wilmsen2013). This unit contains lower Cenomanian ammonites and foraminifera. We have doubts about the presence of the Mantelliceras dixoni Zone fossils at Annopol (see Section 7). The presence of an exclusively M. mantelli Zone benthic foraminifer near the top of unit 4 (see Section 7) suggests that only this ammonite zone of the lower Cenomanian Substage is present at Annopol. In any case, unit 4 matches well the ascending fragment of the early Cenomanian transgressive peak in the British graph (Fig. 10b).

The next tie-points at Annopol for their correlation with the British graph are two closely spaced discontinuities (D4/5 and D5/6 in Fig. 10a), coalescing into a prominent composite hardground discussed by Marcinowski & Walaszczyk (Reference Marcinowski and Walaszczyk1985, fig. 2). A strongly reduced unit 5 is sandwiched between these discontinuities, being practically restricted to the burrow infillings in unit 4 (Fig. 10a). There is a mixture of Turrilites costatus and T. acutus phosphatised ammonites at this level (in addition to the lower Cenomanian forms, also in phosphate preservation) and the foraminifera point to the Rotalipora cushmani Zone. The Cunningtoniceras inerme and its equivalent Thalmanninella reicheli Zone as well as the Acanthoceras jukesbrownei Zone are missing. These missing zones define a minimal extent of gaps embracing unit 5, corresponding to surfaces D 4/5 and D5/6, respectively. These gaps seem to reflect the regressive troughs, whereas the whole unit 5 with its remanié phosphatic fossil assemblages corresponds, albeit in a highly residual form, to the whole middle Cenomanian eustatic low, as seen on the Hancock graph (Fig. 10b; see Hancock, Reference Hancock2004, fig. 2, for detailed presentation and dating of the mid-Cretaceous low).

The middle Cenomanian developments at Annopol are reminiscent of those in the marginal parts of the Anglo-Paris basin. The classic Rouen Fossil Bed replete with phosphatized ammonite moulds may serve here as a close analogue (Juignet & Kennedy, Reference Juignet and Kennedy1976; Hancock, Reference Hancock2004). Local discontinuities D4/5 and D5/6 at Annopol may be tentatively correlated with the Rouen No. 1 Hardground and Rouen No 2. Hardground of the above-mentioned authors, respectively. According to Hancock (Reference Hancock2004), these two hardgrounds correspond in turn to the Primus Event and the Mid-Cenomanian Event of Sussex and the Münsterland (note however that his event definitions do not entirely correspond to those of Ernst, Schmid & Seibertz, Reference Ernst, Schmid and Seibertz1983 and Wilmsen, Reference Wilmsen2003, Reference Wilmsen2013). Also, a glauconite-stained ‘Hg grün’ in the Hannover area, for which Ernst, Schmid & Seibertz (Reference Ernst, Schmid and Seibertz1983) coined the term ‘Mid-Cenomanian Event’, seems to be a closely correlative horizon. All abovementioned levels in Normandy and Germany are associated with the acmes of the irregular echinoid Holaster subglobosus, as is seen at Annopol (see Section 3). In view of the foraminiferal-based correlation of unit 6 with an interval within the Plenus Marl (Metoicoceras geslinianum ammonite Zone) in Eastbourne succession, its seems also probable that an equivalent of the so-called Sub-Plenus Erosion Surface (e.g. Hancock, Reference Hancock1990) may also have merged into the topmost surface of the complex Annopol hardground. If the above correlations are correct, discontinuities D4/5 and D5/6 of the Annopol succession (Fig. 10a) may correspond to sequence boundaries KCe3 and KCe4 of Haq (Reference Haq2014, fig. 2), respectively (with a possibility that his KCe5 is superimposed on KCe4 at Annopol).

The next elements for correlation of the Annopol succession with the British graph are provided with sample 19, representing a lost early late Cenomanian unit and unit 6, dated for the later part of the late Cenomanian (see Section 7). Both these units yield deep-dwelling planktonic foraminifera and are regarded as residual records of the two last transgressive peaks on the Cenomanian portion of the British graph (Fig. 10b). The hardground between the Cenomanian unit 6 and the lower Turonian unit 7 at Annopol (D6/7 in Fig. 10a) may correspond to a regressive trough at the Cenomanian–Turonian boundary (sequence boundary KTu1 of Haq, Reference Haq2014).