Calcareous nannofossil biostratigraphy and paleoenvironmental variations across the upper Paleocene–lowermost Eocene at Gebel Nezzazat, West Central Sinai, Egypt

Detailed lithostratigraphic and calcareous nannofossil biostratigraphic studies were carried out across the Paleocene–Eocene (P-E) that outcrops at Gebel Nezzazat (G. Nezzazat) in West Central Sinai (Egypt). The study interval spans from the upper part of the Tarawan Formation to the lowermost Thebes Formation covering the whole Esna Formation in between them. The Esna Formation had been subdivided into four members: the Hanadi, Dababiya Quarry, Mahmiya, and Abu Had Members. Five calcareous nannofossil biozones (NP7/8, NP9 to NP12) and four subzones (NP9a and NP9b and NP10a and NP10b) were recognized. The lowest occurrences (LOs) of Fasciculithus alanii group, Neochiastozygus junctus, Sphenolithus radians, and Blackites herculesii as well as the highest occurrence of F. alanii group and the increased frequency of N. junctus are biostratigraphically significant. On contrast, the LOs of Discoaster binodosus, Discoaster mahmoudii, Discoaster diastypus, Zygrhablithus bijugatus, and Campylosphaera dela as well as the LOs of Fasciculithus tympaniformis are unreliable bioevents. Calcareous nannofossils increase in abundance close to the P-E transition. Ericsonia subpertusa suddenly increases above the base of Eocene, whereas the diversity of Fasciculithus drops close to this level. The P-E boundary at G. Nezzazat was placed at the base of the Dababiya Quarry Member in coincidence with the base of Subzone NP9b that was delineated by the LOs of Discoaster araneus, Rhomboaster cuspis, Rhomboaster calcitrapa, Rhomboaster spineus, and Rhomboaster bitrifida. A small gap was recorded across the P-E boundary as indicated by the lack of the four beds of the Dababiya Quarry Member. The changes in calcareous nannofossil assemblages reveal warm-water and oligotrophic conditions prevailed during the transition at G. Nezzazat.


Introduction
The Paleocene-Eocene was distinguished by various climatic, sedimentologic, isotopic, and paleontologic variations . This interval was accompanied by a short-lived episode of global warming associated with negative shifts in the δ 13 C and δ 18 O values as well as carbonate

Responsible Editor: Zakaria Hamimi
This article is part of the Topical Collection on Current Advances in Geological Research of Egypt contents (e.g., Cramer et al. 2003;Zachos et al. 2005;Westerhold et al. 2009;Frieling et al. 2018). Furthermore, the P-E transition was associated with changes in the biotic assemblages which include originations of new species, radiations, migration towards higher latitudes, and turnovers of some taxa (e.g., Bralower et al. 2002;Gingerich 2003;Bown et al. 2004;Tantawy 2006;Agnini et al. 2007a;Westerhold et al. 2009;Kasem et al. 2020a, b). As a result of the location of Egypt at the Tethys' southern margin during the P-E transition, Egypt had complete P-E sequences, which have a perfect record of changes in sedimentology and biotic assemblages throughout this interval Kasem et al. 2020b).
Thus, Egyptian sequences attracted several authors to track the biotic, isotopic, and climatic responses to climatic variations through the P-E interval (e.g., Youssef 2016;Obaidalla et al. 2017;Faris et al. 2018;Kasem et al. 2020a, b). The Global Boundary Stratotype Section and Point (GSSP) that hosts the base of the Eocene Series had been officially ratified at the Dababiya Quarry in southern Egypt . The onset of the global carbon isotope excursion (CIE) had been regarded as the most useful characteristic that marks the base of the Eocene Series at the GSSP . The base of Eocene at the GSSP coincides with a distinctive bed known as the Dababiya Quarry Beds (DQBs) that are easily distinguished based on their lithological, biotic, and isotopic features . Further subdivisions to the DQBs were achieved by several authors based on the variations in color, texture, and the phosphatic as well as carbonate contents (e.g., Aubry et al. 2007;Ouda et al. 2013). The DQBs had been recorded in many sections in Egypt (e.g., Tantawy 2006a, b;Ouda et al. 2016a, b;Faris et al. 2017;Kasem et al. 2020b).
Calcareous nannofossils are highly sensitive to climatic changes and underwent considerable variations during the late Paleocene-early Eocene. Therefore, they are a significant tool for tracking the biotic response across this transition (e.g., Perch-Nielsen 1985;Bralower et al. 2002;Dupuis et al. 2003;Tantawy 2006a, b;Agnini et al. 2006, Agnini et al. 2007aBernaola et al. 2007;Mutterlose et al. 2007; Raffi and Raffi and De Bernaola 2008;Bown and Pearson 2009;Raffi et al. 2009;Self-Trail et al. 2012;Kasem et al. 2020a, b). The calcareous nannofossil assemblage's changes  that include the origination and radiation of genus Rhomboaster and the origination of new species of genera Discoaster and Fasciculithus are important bioevents that mark the P-E boundary, and are useful in global biostratigraphic correlations (Table 1).
The main intents of this study are to describe the lithostratigraphy of the uppermost Paleocene-lowermost Eocene succession at Gebel Nezzazat, to provide detailed calcareous nannofossil biostratigraphy, to reveal the variations in calcareous nannofossil assemblages, and to discuss the biostratigraphic reliability of significant bioevents across the P-E interval, as well as to track the changes in water temperature and fertility through the P-E transition.

Location, material, and methods
Sixty-one rock samples covering the Paleocene-Eocene sequence at Gebel Nezzazat (Lat. 28° 47′ 45.4″N and Long. 33° 14′ 9.79″E), West Central Sinai, Egypt (Fig. 1), were investigated in this study. Smear slides from each sample were prepared for investigation of calcareous nannofossil assemblages according to Bown and Young (1998). The slides had been investigated by a Euromex iScope microscope at 1250× magnification at Damanhour University, Egypt. The state of preservation of the calcareous nannofossil assemblages is good, where little or no evidence of overgrowth and/or etching was noted. A quantitative analysis was carried out to calcareous nannofossil assemblages. Specimens had been counted in 50 fields of view to reveal the abundances of various species. Additional 50 fields of view were viewed just to reveal scarce species. To ensure even distribution of the sediments on the slide, about 0.1 g of raw sediments was transferred into 10 ml of distilled water in a test tube and they were dispersed; 0.25 mm of the suspension was pipetted onto 22 mm × 22 mm coverslips and dried on a hot plate. The coverslips were fixed onto glassy slides by using DPX Mountant. The materials used for this study are deposited at Faculty of Science, Geology Department, Damanhour University, Damanhour, Egypt.  s  u  i  d  e  m  r  e  t  n  i  s  u  h  t  i  l  o  c  a  l  p  i  c  u  r  C   s  o  t  s  a  g  e  t  s  s  u  i  h  c  y  t  p  o  h  t  i  L   s  i  s  n  e  i  d  a  b  r  a  b  r  e  t  s  a  o  c  s

Lithostratigraphy
The uppermost Paleocene-lowermost Eocene at Gebel Nezzazat (G. Nezzazat) extends in three formations that are, from base to top, the Tarawan, Esna, as well as Thebes Formations (Fig. 3, Table 2). Awad and Ghobrial (1965) had introduced this unit to describe a succession of chalk passing into chalky limestones and marly limestone at Gebel Tarawan, Kharga Oasis (Western Desert, Egypt). About 2.3 m from the upper part of the Tarawan Formation had been investigated in this study and consists of hard, yellow argillaceous limestone grading into white chalky limestone (Figs. 3 and 4A, B). It lies conformably above the Dakhla Formation and below the Esna Formation (Fig. 4A). Depending on the calcareous nannofossils, it was assigned to the (NP7/8 Zone) Thanetian Age (Figs. 3 and 5).

The Esna Formation
The Esna Shale Passage Beds was originally established by Beadnell (1905) to characterize a succession of 104-m-thick greenish gray shale intercalated with calcareous interbeds, grades upward into argillaceous limestone, and overlies the Cretaceous sequence and underlies the Eocene succession at Gebel Oweina, near Esna, southern Egypt. He named it Esna Shale. This unit was assigned to the Esna Formation by Said (1960). At G. Nezzazat, this unit is about 9.2 m of gray to dark gray shale intercalated with black shale and yellow, gray to brown calcareous shale, and upgrades to yellow marl and white chalky limestone (Figs. 3 and 4A, C, D). It conformably overlies the Tarawan Formation and is overlain by the Thebes Formation conformably (Fig. 4A, C, D). The Esna Formation in our study contains the following calcareous nannofossil biozones (the upper part of NP7/8, NP9, NP10, and NP11 and the lower part of NP12), so it is assigned to Thanetian-Ypresian ages (Figs. 3 and 5).

The Hanadi Member
This unit was originally established to extend from the ground of the Esna Formation (Fm) to the top of the phosphatic bed (Abdel Razik 1972). Later, Aubry et al. (2007) delimit this unit to extend from the base of the Esna Fm to the base of the phosphatic bed. It consists mainly of gray, massive, calcareous shales. At G. Nezzazat, this member consists of 1.8-m-thick gray to dark gray shale (Fig. 3).

The Dababiya Quarry Member
It was originally established by Aubry et al. (2007) to distinguish the distinct beds that mark the base of the Eocene at the Dababiya Quarry area in Egypt. These marker beds reflect climatic changes in the southern Tethys during the P-E transition as indicated by their lithological, geochemical, and biotic characteristics (Aubry et al. 2007). The base of this unit in this study is coincident with the entry of Subzone NP9b (Fig. 3) Kasem et al. 2020b). Further subdivision of these beds was introduced by several authors depending on their color, texture, contents of carbonate and coprolite, and/or the composition and patterns of foraminifers (e.g., Aubry et al. 2007;Ouda et al. 2013). Moreover, Tantawy (Tantawy 2006a, b) subdivided the interval of the Esna Formation that covers the Dababiya Quarry Member of Aubry et al. (2007) into five units (DQBs 1-5). The DQB1, at section E, which lies east of the P-E GSSP is missed (Khozyem et al. 2015). At Naqb Assiut section, Kharga Oasis, Western Desert, Egypt, the Dababiya Quarry Member is only represented by the upper two beds (nos. 4 and 5) (El-Dawy et al. 2016), while at Gaga section, Baris Oasis, Western Desert, Egypt, the first bed (clay layer) of the Dababiya Quarry Member at the Paleocene/Eocene GSSP which represents the onset of the P-E boundary is missing, whereas the rest of the beds (nos. 2-4) are well represented (Metwally and Mahfouz 2018). In the present study, the Dababiya Quarry Member (DQM) is represented by only one sample of black shale, which may represent DQB1, of about 15 cm thick (Fig. 3). The Mahmiya Member Aubry et al. (2007) established this member to cover the main part of the Esna Unit 2 of Dupuis et al. (2003) at the Dababiya Quarry section. It consists of monotonous, dark, clayey shales without marked bedding and has low calcium carbonate content (< 50%) and clear cyclic color variations. The thickness of the Mahmiya Member is 65 m in the Dababiya Quarry section. Aubry et al. (2007) used the term El Quda Bed to describe the thin (10-30 m) calcarenite bed within the Mahmiya Member which consists generally of accumulations of cylindrical coprolites, phosphatic and shell clasts, and a variable amount of glauconite at the base (Table 2). It differs from the beds of the DQM by the presence of glauconite and the erosive and bioturbated contact with the underling lithology. In this study, the Mahmiya Member is ~ 3.4 m of gray to dark gray shales (Fig. 3).

The Abu Had Member
This member was originally established by Abdel Razik (1972) as the lower member of the Thebes Formation and covers the interbeds of limestones and shales at the transition between the shales of the Esna Formation and the massive carbonates of the Thebes Formation. Aubry et al. (2007) assigned this member to the Esna Formation at the Dababiya Quarry (Table 2). At the Dababiya Quarry, this member is 43.5 m thick of alternating shales and limestone. In the present study, the upper fraction of the Esna Formation at the G. Nezzazat section is assigned to the Abu Had Member and is about 4 m of yellow calcareous shale, gray shale, gray to brownish calcareous shale, dark gray shale, yellow to gray calcareous shale, yellow marl, white chalky limestone associated with chert nodules and bands, as well as brown calcareous shale (Fig. 3).

The Thebes Formation
It was described by Said (1960) to cover a 290-m white to grayish white limestone with chert bands that superimpose the Esna Formation at the Gebel Gurnah, southern Egypt. About 2 m of hard yellow to white argillaceous limestone from the lowermost part of this formation was examined at G. Nezzazat and was assigned to the Ypresian Age based on the presence of the calcareous nannofossil biozone (NP12) (Figs. 3, 4A, D, and 5). It conformably underlain by the Esna Formation. Martini's (1971) zonation scheme of the Paleogene was adopted in this study, and Romein's (1979) recommendation of combining Zones NP7 and NP8 was followed. In addition, Aubry et al.'s (1999) proposition of subdividing Zone NP9 into Subzones NP9a and NP9b was applied. Furthermore, the subdivision of Zone NP10 suggested by Kasem et al. 2020a into Subzones NP10a and NP10b was followed. The stratigraphic importance of calcareous nannofossil bioevents that associate the P-E transition was discussed in this study. Abbreviations that are applied in this study are LO for the lowest occurrence, HO for the highest occurrence, LCO for the lowest common occurrence, and HCO for the highest common occurrence (Kasem et al. 2017b). Table 1 shows the counts and stratigraphic distributions of calcareous nannofossil assemblages. Microphotographs of the recorded species are shown in Plates 1 and 2. Five calcareous nannofossil zones and four subzones were recognized and discussed as follows.

Discoaster mohleri Zone (NP7/8)
This combined zone spans from the LO of taxon Discoaster mohleri to the LO of Discoaster multiradiatus (Romein 1979). It is comparable to Zones CP6 and CP7 of Okada and Bukry (1980) and Zones CNP9 and CNP10 of Agnini et al. (2014). Originally, it was introduced by Hay (1964), yet the  marker species Heliolithus riedelii was not recorded from several sections besides having inconsistent stratigraphic ranges (Perch-Nielsen 1985;Romein 1979;Agnini et al. 2007b). Accordingly, Romein (1979) excluded this bioevent from being a zonal marker and lumped Zones NP7 and NP8 together into the D. mohleri Zone (NP7/8). In this study, this zone is ~ 2.8 m thick (samples 1 to 11) and covers almost the whole Tarawan Formation and extends to the basal part of the Esna Formation (Table 1, Fig. 2). It was assigned to Thanetian.

Discoaster multiradiatus Zone (NP9)
This zone had been established by Bramlette and Sullivan (1961) and then had been emended by Martini (1971) to the biostratigraphic interval between the LOs of Discoaster multiradiatus and Tribrachiatus bramlettei. It is comparable to Zone CN8 of Okada and Bukry (1980), Subzone NTp16b to Zone NTp20 of Varol (1989), and Zone CNP11 plus the lower part of Zone CNE1 of Agnini et al. (2014). Zone NP9 was documented from several Paleocene-Eocene successions in Egypt (Kasem et al. 2020a, b). At G. Nezzazat, this zone extends from sample 11 to sample 22 and is ~ 2.4 m thick in the basal portion of the Esna Formation and dates Thanetian-Ypresian (Table 1, Fig. 3).

NP9a Subzone
It covers the biostratigraphic interval between the LOs of Discoaster multiradiatus and Rhomboaster spp., D. araneus, plus/or D. anartios . It extends from sample 11 to sample 18 at G. Nezzazat (Table 1) and covers about 1.5 m in the lower portion of the Esna Formation (Fig. 3). This subzone belongs to the Thanetian Age. Several new species come in this interval; among them are Fasciculithus lillianiae, Neochiastozygus pusillus, Ellipsolithus distichus, Discoaster barbadiensis, Discoaster falcatus, Neochiastozygus distentus, and Lanternithus simplex (Table 1).

NP9b Subzone
The base of this subzone was located by the LOs of Rhomboaster spp., D. araneus, and/or D. anartios, and its upper limit was delineated by the LO of taxon Tribrachiatus bramlettei . It is 0.9 m thick in the Esna Fm at G. Nezzazat (Fig. 3) (Faris and Salem 2007). Some authors documented that Fasciculithus alanii go extinct in coincidence with the ground of Subzone NP9b (Faris and Salem 2007), yet it has its HO earlier at the Dababiya Quarry , and extend up to the lowermost Eocene in some sections (Kasem et al. 2020b). In this study, F. alanii occurs in Subzone NP9b (Table 1).
1 3 (1980) and the upper part of Zone CNE1 and Zone CNE2 of Agnini et al. (2014). It extends from sample 23 to sample 34 (~ 2.3 m thick) in the Esna Fm (Fig. 3) and is assigned to the Ypresian Age.
The start of Zone NP10 was delineated variously by different authors because of the taxonomic disputes among them concerning whether T. bramlettei has a different structure from Rhomboaster or not (Kasem et al. 2020a).  proposed to locate the base of Zone NP10 at the LCO of T. bramlettei. At G. Nezzazat, T. bramlettei first occurs in sample number 23 with common occurrence (Table 1) and has been adopted to mark the base of Zone NP10. Abu  suggested approximating the base of Zone NP10 at the increased frequency of Neochiastozygus junctus to delineate the lower limit of Zone NP10. At G. Nezzazat, N. junctus first occurs in sample 19 and shows increased frequency in sample 21 slightly below the base of Zone NP10 (sample 23) ( Table 1).
The genus Fasciculithus commonly disappears in the basal part of Zone NP10 (Romein 1979); thus, it can help in the approximation of the lower limit of this zone in case of the scarcity or absence of the zonal marker (Perch-Nielsen 1985). However, Kasem et al. (2020a) noted that F. tympaniformis occurs with common occurrence up to the middle part of Zone NP10. Representatives of genus Tribrachiatus are important in biostratigraphic correlation and further subdivision of Zone NP10. The co-occurrences of Tribrachiatus bramlettei and T. contortus in Zone NP10 from one side and T. contortus-Tribrachiatus orthostylus from another side are a good indicator to the completeness of this interval (e.g., Agnini et al. 2006;Kasem et al. 2020a, b).
At G. Nezzazat, an overlap between the stratigraphic ranges T. bramlettei and T. contortus which were documented extends from sample 26 to sample 33, and T. contortus and T. orthostylus have a stratigraphic overlap extending from sample 26 to sample 34 (Table 1). Some authors suggested further subdivisions for Martini's (1971) Zone NP10 (e.g., Aubry 1996; Kasem et al. 2020a). Aubry (1996) suggested subdividing Zone NP10 into four subzones: NP10a to NP10d, based on LOs and HOs of Tribrachiatus digitalis and T. contortus. The validity of this suggestion was confirmed by various studies (e.g., Dupuis et al. 2003;Al Wosabi 2015). Yet the reliability of T. digitalis as a marker is questionable because of the disputes concerning its taxonomic position and biostratigraphic range . Consequently, Kasem et al. (2020a, b) recommended to exclude T. digitalis from being a reliable subzonal marker and used the LO T. contortus to subdivide Zone NP10 into Subzone NP10a plus NP10b. This suggestion is followed in this study. Kasem et al. (2020a) introduced Subzone NP10a to extend between the subsequent LOs of Tribrachiatus bramlettei and T. contortus. This subzone is comparable to Aubry's (1996) Subzones NP10a to NP10c and Tantawy's (1998) Subzones NP10a and NP10b. At G. Nezzazat, this subzone extends from sample 23 to sample 25 (~ 40 cm thick) in the Esna Formation (Fig. 3, Table 1) and is assigned to the Ypresian Age.

Discoaster binodosus Zone (NP11)
This zone covers from the HO of Tribrachiatus contortus to the LO of Discoaster lodoensis as delineated by Hay and Mohler (1967). It is comparable to Subzone CP9b of Okada and Bukry (1980) and is equivalent to Zone CNE3 of Agnini et al. (2014). At G. Nezzazat, this zone is ~ 3.9 m thick in the Esna Formation (Table 1, Fig. 3) and is assigned to the Ypresian Age.

3
Tribrachiatus orthostylus commonly appear close to the base of Zone NP11 and, therefore, can approximate the entry of this zone when its zonal marker is absent (Perch-Nielsen 1985). In this study, T. orthostylus appears with sporadic occurrence at the base of Subzone NP10b; however, its lowest continuous occurrence (LCtO) and increased frequency are coincident with the base of Zone NP11 (sample 35) (Table 1). Moreover, the LCtO and increased frequency of T. orthostylus are coincident with the LO of Sphenolithus radians (Table 1) and can approximate the entry of Zone NP11 when T. contortus is absent Kasem et al. 2020a).

Tribrachiatus orthostylus Zone (NP12)
It ranges from the LO of Discoaster lodoensis to the HO of taxon Tribrachiatus orthostylus (Brönnimann and Stradner 1960). It is correlative to Okada and Bukry's (1980) Zone CP10 and Zone CNE4 of Agnini et al. (2014). It covers ~ 2 m from the uppermost Esna Fm to the lowermost Thebes Fm and is assigned to the Ypresian Age (Table 1, Fig. 3). In case of poor preservation of the zonal marker, the base of Zone NP12 can be placed approximately based on the LOs of Sphenolithus conspicuous that often first occurs in the topmost of Zone NP11 and the LO of Neococcolithus dubius that commonly first occurs in the lowermost of Zone NP12 (Abu ).

Discussion
The latest Paleocene-earliest Eocene interval was characterized by biotic and lithological changes . These variations are important in limiting the P-E boundary and biostratigraphic purposes (Bukry 1973;Aubry 1996;Kasem et al. 2020a, b, etc.).

Calcareous nannofossil bioevents
Considerable changes in calcareous nannofossils are associating the Paleocene-Eocene Thermal Maximum (PETM) and are useful for the placement of the P-E boundary (Romein 1979;Dupuis et al. 2003;Raffi et al. 2005;Agnini et al. 2007a, b). These changes include incoming of new species, radiations, migration towards higher latitudes, and turnovers of some taxa (Kasem et al. 2020a, b).

The LO and HO of Discoaster multiradiatus
Discoaster multiradiatus had been regarded as a reliable Paleocene zonal marker by various authors (e.g., Varol 1989;Agnini et al. 2017). Moreover, this species is supposed to go extinct in the lower portion of Zone NP11 (Perch-Nielsen 1985); however, Kasem et al. (2020a) noted that it disappears in Subzone NP10b. At Nezzazat, it disappears within the upper part of Zone NP11 (Table 1). Thus, the HO of D. multiradiatus seems to be inconsistent and unreliable for biostratigraphic correlations.

The LO of the Fasciculithus alanii group
The last radiation of Fasciculithus occurs during the latest Paleocene (Romein 1979). Previous studies revealed that F. alanii first occurs in Subzone NP9a and disappears in this subzone close to the CIE Faris and Farouk 2015). Thus, it can distinguish the Paleocene portion of Zone NP9 from the Eocene part (Aubry and Salem 2013a). However, Faris and Abu Shama (2007) recorded F. alanii shortly above the base of Subzone NP9b. Agnini et al. (2007b) lumped Fasciculithus richardii, F. mitreus, F. hayi, and Fasciculithus schaubii within the F. richardii group. Similarly, Agnini et al. (2014Agnini et al. ( , 2017 lumped Fasciculithus species that first appear in Zone NP9 into the F. richardii group. Later, Kasem et al. (2020a) recommended assigning this group to the F. alanii group, where the occurrence and disappearance of F. alanii were the most important bioevents in this interval (Agnini et al. 2007a, b;Aubry and Salem 2013a).
At G. Nezzazat, F. richardii appeared in sample 10 just below the base of Zone NP9, whereas F. thomasii appears in sample 10 in coincidence with the base of Zone NP9 and followed by the appearance of F. alanii in sample 12 then F. lillianiae in sample 13 (Table 1). These taxa were grouped in the F. alanii group and seem to be useful in the approximation of the base of Zone NP9 when its zonal marker is absent or poorly preserved.

The LOs and HOs of the calcareous nannofossil excursion taxa
Calcareous nannofossil excursion taxa (CNET) or the Rhomboaster-Discoaster (RD) assemblage includes Discoaster araneus and D. anartios, as well as Rhomboaster spp. They appear suddenly and dominate the nannofossil assemblages of the lowermost Eocene Kahn and Aubry 2004;Bown and Pearson 2009). Kahn and Aubry (2004) included D. falcatus, Bomolithus supremus, Coccolithus bownii, and Toweius serotinus in the CNET.

3
The CNET which had been considered the most remarkable change in calcareous nannofossils marks the P-E boundary and points to unusual conditions prevailed during this interval ; Table 3).
The LOs of the CNET were useful for further subdivision of Zone NP9 (Aubry and Sanfilippo 1999;Agnini et al. 2007a, b;Kasem et al. 2020a, b). Yet, the lowermost Eocene (~ 73 cm) is marked by a drop in carbonate content and, subsequently, the dissolution of calcareous plankton at the GSSP , which make it hard to either recognize the exact LOs of the CNET or track the precise changes in calcareous nannofossil assemblages throughout the P-E transition .
At G. Nezzazat, the LOs of Rhomboaster calcitrapa, R. cuspis, Rhomboaster spinosa, and Rhomboaster bitrifidia Discoaster araneus were used to differentiate the Paleocene Subzone NP9a from the Eocene Subzone NP9b (Table 1, Fig. 3). Furthermore, it was suggested to assign the interval from the HO of CNET and the top of Zone NP9 to Subzone NP9c (Aubry and Salem 2013a). In the present study, the CNET extends up the upper part of Zone NP10 (Table 1).

The LO of Discoaster mahmoudii
Previous studies revealed that Discoaster mahmoudii appears in Zone NP9 Aubry and Salem 2013b). Aubry and Salem (2013b) pointed to the restriction of the LO of D. mahmoudii to the top section of Zone NP9 (Subzone NP9c). However, it appears to be ~ 3.5 m above the base of Eocene at the Dababiya Quarry . Moreover, it was recorded within Zone NP10 at ~ 7 m above the CIE at G. Nezzi .
At G. Nezzazat, the LO of D. mahmoudii is coincident with the base of Subzone NP9b (Table 1). Thus, the biostratigraphic significance of this event is weakened (Agnini et al. 2007a;Kasem et al. 2020a).

The LO of Blackites herculesii
This species had been introduced as Rhabdosphaera herculea (Stradner 1969) then emended to Rhabdolithus solus (Perch-Nielsen 1971) and later revised to Blackites herculesii (Bybell and Self-Trail 1997). It had been documented in Zone NP9 (Kasem et al. 2020a, b) and was suggested to be restrictive to the topmost of this zone (Aubry and Salem 2013a). However, it was recorded in Zone NP10 at the Misheiti section in Sinai, Egypt (Kasem et al. 2020a).
At G. Nezzazat, it appears in the topmost portion of Zone NP9b (sample 21) (Table 1). Thus, it can approximate the top of Zone NP9.

The LO of Campylosphaera dela
It is hard to distinguish between Campylosphaera dela and C. eodela (Bramlette and Sullivan 1961;Hay and Mohler 1967;Bukry and Percival 1971). Therefore, they were regarded as synonyms (Kasem et al. 2020a, b). Bukry (1973) used the LO of C. dela in subdividing the interval equivalent to Martini's (1971) Zone NP9. However, Kasem et al. (2020a) noted this species in Zone NP7/8. Thus, the stratigraphic range of this taxon is inconsistent (Agnini et al. 2007b).
In this study, C. dela and C. eodela were viewed as a single taxon and first occurred in the topmost of NP9a (sample 18) just below the base of Subzone NP9b (Table 1). As a result of the inconsistent ranges and taxonomic disputes, C. dela was excluded from being a reliable biostratigraphic event (Kasem et al. 2020a).

The LO of Discoaster binodosus
It had been proposed to use the LO of Discoaster binodosus in the approximation of the top of Zone NP9 when its zonal marker, Tribrachiatus bramlettei, is scarce or absent (Perch-Nielsen 1985); however, the range of this taxon is inconsistent where it was recorded in the lower part of Zone NP9 (Faris and Salem 2007) and within Zone NP10 (Kasem et al. 2020a, b). Thus, the biostratigraphic significance of these bioevent is weakened (Faris and Salem 2007;Kasem et al. 2020a, b).
In this study, D. binodosus appears within Subzone NP9b, but it is sporadic in the early stage of its range and its LCO is coincident with the base of Subzone NP10b (Table 1).

The drop in the frequency and diversity of Fasciculithus species
Several studies documented a severe decrease in the diversity of Fasciculithus close to the upper limit of Paleocene (e.g., Agnini et al. 2007a, b;Kasem et al. 2020a).
In this study, the diversity of Fasciculithus decreases from 10 species in sample 17 slightly below the ground of Subzone NP9b into 5 and 3 just below and above the P-E boundary at G. Nezzazat (Table 1, Fig. 7). Furthermore, the abundance of Fasciculithus/Lithoptychius drops from 179 specimens per 50 fields of view (S/50 FOV) in sample 17 into 51 and 11S/50 FOV specimens just below and above the base of the Eocene (Table 1, Fig. 7).

The HOs of Fasciculithus tympaniformis and F. alanii group
Fasciculithus taxa diversified in the uppermost Paleocene and go extinct shortly above this interval (Martini 1971;Raffi et al. 2005;Tantawy 2006a, b). Thus, it can approximately locate the top of Zone NP9 when the zonal marker Tribrachiatus bramlettei is poorly preserved or absent (Perch-Nielsen 1985). Agnini et al. (2014Agnini et al. ( , 2017 used the HO of F. tympaniformis to mark the upper limit of the earliest Eocene Zone CNE1. However, several Fasciculithus species were recorded up to Zone NP10 (Romein 1979;Bralower and Mutterlose 1995;Aubry 1996;Raffi et al. 2005;Agnini et al. 2007b). Kasem et al. (2020a) noted that F. tympaniformis extends with common occurrences up to the topmost section of Zone NP10. In this study, F. clinatus, F. involutus, Fasciculithus billii, Fasciculithus pileatus, Fasciculithus janii, and Fasciculithus bitectus disappear in Zone NP9 (Table 1). However, F. tympaniformis occurs with common occurrences up to Zone NP9 and extends with rare and discontinuous occurrences up to Zone NP11 (Table 1), which indicates reworking of these specimens and supports the biostratigraphic reliability of its HCO rather than the HO that is an inconsistent bioevent, possibly due to the diachronous nature of the HO of this taxon or discrepancies concerning the taxonomic position of T. bramlettei (Tantawy 2006a, b;Kasem et al. 2020a). Agnini et al. (2014Agnini et al. ( , 2017 documented the HO of F. richardii group below the LOs of Rhomboaster spp. and suggested delineating the base of the entry of the Eocene (base of Zone CNE1) based on the HO of this group. Abu  recorded the HO of F. alanii in coincidence with the base of Subzone NP9b. Thus, the HO of F. alanii group can approximate the lower limit of Subzone NP9b when its subzonal markers are absent or poorly preserved.
In this study, the HOs of F. lillianiae and F. thomasii are coincident with the base of Subzone NP9b, whereas F. alanii extends within Subzone NP9b (Table 1).

The acme of Ericsonia subpertusa
An increased frequency of Ericsonia subpertusa was documented across the P-E interval Abu Shama et al. 2007;Kasem et al. 2020a).
At G. Nezzazat, E. subpertusa dominates the latest Paleocene-earliest Eocene nannofossil assemblages; nevertheless, it suddenly increases from 121 S/50 FOV in samples 17 to 270 and 446 S/50 FOV just below and above the P-E boundary ( Table 1).

The LO of Discoaster diastypus
The appearances of both Tribrachiatus contortus plus Discoaster diastypus had been regarded synchronous . Therefore, Bukry (1973) used their LOs to limit the base of the Discoaster diastypus Zone of Bukry (1973) that is correlative to Zone CP9 of Okada and Bukry (1980). The co-occurrences of T. bramlettei and D. diastypus were recorded in some areas in Egypt (e.g., Abu ), yet their LOs were not concurrent in other sections (Tantawy 2006a, b;Kasem et al. 2020a).
At G. Nezzazat, the LO of D. diastypus precedes the appearance of T. bramlettei (Table 1) that supports the diachronous nature of D. diastypus and/or T. bramlettei.

The LO of Sphenolithus radians
The appearance of Sphenolithus radians and the disappearance of species Tribrachiatus contortus were very associated; therefore, the LO of S. radians can approximate the entry of Zone NP11 when T. contortus is absent (Perch-Nielsen 1985). Kasem et al. (2020a) noted that the LOs of S. radians and T. contortus are coincident at Misheiti in Sinai, Egypt. However, S. radians were recorded up to the topmost of Zone NP11 (Abu Al Wosabi 2015).
AT G. Nezzazat, the LO of S. radians and the HO of T. contortus are coincident (Table 1), which supports the reliability of this bioevents in approximation of the base of Zone NP11.

The LO of Zygrhablithus bijugatus
The LO of Zygrhablithus bijugatus at the base of Subzone NP9b was recorded (Kasem et al. 2020b). In this study, it occurs at the base of Subzone NP10b (Table 1). This indicates inconsistent LO of this taxon. Kasem et al. (2020b) noted that Neochiastozygus junctus first occurs in Subzone NP9b and blooms just above the top of Zone NP9. The increased frequency of this taxon was used to limit the basis of Zone NP10 (Abu .

The LO and increased frequency of Neochiastozygus junctus
At G. Nezzazat, the LO of N. junctus is coincident with the ground of Zone NP9b and increases in frequency in the upper section of Zone NP9 and in the lower part of Zone NP10 in this study (Table 1). Thus, this bioevent can approximate this interval in case of the absence of the zonal markers (Table 1).

The LOs and HOs of Tribrachiatus species
Genus Tribrachiatus is useful in biostratigraphic zonation (Martini 1971). Previous investigations documented that T. bramlettei appears slightly above the base of the Eocene (Agnini et al. 2007a, b). Yet its earliest occurrences are rare to very rare and sporadic (Agnini et al. 2007a).
Thus,  suggested depending on the LCO of this taxon to delineate the lower limit of Zone NP10.
Furthermore, the LO of T. bramlettei is inconsistent (Agnini et al. 2007b), probably due to the diachronous nature of this bioevent and/or dissolution that affects the lowermost Eocene sediments (Agnini et al. 2007b). In addition, T. bramlettei and Rhomboaster cuspis were considered synonyms by some authors (e.g., Von Salis et al. 2000); whereas others count them two different species (e.g., Raffi et al. 2005;Kasem et al. 2020a, b; the present study).
Furthermore, several authors reported that Tribrachiatus bramlettei disappears in the uppermost of Zone NP10, shortly after the HO of T. contortus and below the LO of T. orthostylus (e.g., Aubry and Sanfilippo 1999). Nevertheless, T. bramlettei extends up to the top of Zone NP10 at Galala Mountains, Eastern Desert, Egypt (Marzouk and Scheibner 2003).
In this study, the HO of T. bramlettei occurs in sample 33 at a level somewhat below the end of Zone NP10 and slightly below the LCtO of T. orthostylus (Table 1).
The taxonomic position of Tribrachiatus digitalis underwent criticism , and consequently, its biostratigraphic significance is weakened (Kasem et al. 2020a, b). On the other hand, T. contortus is significant in biostratigraphic studies (Martini 1971), where its HO delineates the top of Zone NP10, and its LO is useful in subdividing Zone NP10 into Subzones NP10a and NP10b (Kasem et al. 2020a, b, and the present study). The appearance of T. orthostylus can approximate the topmost of Zone NP10 when its zonal marker is absent (Perch-Nielsen 1985). In this study, T. orthostylus appears just below the base of Subzone NP10b (Table 1).

The calcareous nannofossil species richness and abundance
Calcareous nannofossils reached their maximum diversity in the Paleogene interval within Zone NP9 (Perch-Nielsen 1985 Fig. 6).

Paleoecology
The changes in temperature and nutrient availability affect the distribution of calcareous nannofossil; therefore, they can be used for tracking climatic changes (e.g., Bernaola et al. 2007). Oligotrophic conditions with warm waters that characterize low-to middle-latitude areas are sufficient for flourishing modern calcareous nannofossil assemblages (e.g., Watkins 1989;Bornemann 2003;Fuqua et al. 2008).
On the other hand, eutrophic forms are more common in cold nutrient-rich waters at high-latitude areas . Several investigations revealed the paleoecological preferences of several calcareous nannofossil taxa (e.g., Wei and Wise Jr 1990;Fuqua et al. 2008, etc.).

Remarks
The Paleocene-Eocene transition was marked by geochemical, lithological, and biotic variations . Based on calcareous nannofossils, the P-E boundary had been placed either in Zone NP9, at the NP9/10 zonal limit, or in Zone NP10 (Martini 1971;Bukry 1973;Bolle et al. 2000). At the Dababiya Quarry section in Egypt that was ratified as the GSSP for the base of Eocene, the P-E boundary was placed at the ground of Subzone NP9b ). This boundary is marked by an abrupt decrease in carbon isotope and is coincident with the base of a distinctive bed of the Dababiya Quarry Member (Aubry et al. 2007).
At G. Nezzazat, this boundary had been placed at the NP9a/NP9b subzonal boundary that is denoted by the appearances of Discoaster araneus and/or Rhomboaster spp. (Table 1). This boundary is marked by the LOs of Rhomboaster cuspis, R. calcitrapa, R. spineus, R. bitrifida, D. araneus, and Chiasmolithus nitidus as well as the HOs of Fasciculithus thomasii and F. lillianiae (Table 1). In addition, this level is coincident with the base of a black shale bed assigned to the DQM (correlative to DQB 4 of Tantawy 2006, ) in the lower portion of the Esna Formation (Fig. 3).
This reveals the presence of a gap represented by the missing of DQB2 to DQB5, but not detected by the calcareous nannofossils. Similar results were noted in different sections in Sinai and others in Egypt (see Aubry and Salem 2013b;Obaidalla et al. 2017;Kasem et al. 2020a, b for more discussion). Variations in calcareous nannofossil assemblages reveal warming and oligotrophic conditions prevailed during the P-E transition at G. Nezzazat in agreement with results from the GSSP as well as sections from various areas in the world Bernaola et al. 2007;Mutterlose et al. 2007;Raffi et al. 2009;Kasem et al. 2020a, b).

Conclusions
Lithostratigraphic and calcareous nannofossil biostratigraphic investigations were carried out in this study at the Gebel Nezzazat section in Central Sinai, Egypt. The interval investigated extends throughout the Tarawan, Esna, and Thebes Formations. The Esna Formation has been partitioned into the Hanadi, Dababiya Quarry, Mahmiya, and Abu Had Members.
Five calcareous nannofossil biozones were recognized (NP7/8 through NP12). Zone NP9 had been subdivided into Subzones NP9a and NP9b depending on the LOs of Rhomboaster spp. and/or Discoaster araneus. Moreover, the LO of T. contortus had been used to partition Zone NP10 into Subzones NP10a and NP10b.
Exclusion of Tribrachiatus digitalis from being a reliable marker was recommended in this study. Stratigraphic overlaps between T. bramlettei and T. contortus as well as between T. contortus and T. orthostylus were noted, indicating the completeness of this interval. Most Fasciculithus taxa (F. clinatus, F. involutus, F. billii, F. pileatus, F. janii, and F. bitectus) disappear within Zone NP9. Fasciculithus tympaniformis occurs with common occurrences up to the top of Zone NP9 and extends with sporadic occurrences up to Zone NP11 (Table 1), supporting the insignificance of the HO of F. tympaniformis in biostratigraphy.
Subsequent appearances of F. richardii, F. thomasii, F. alanii, and F. lillianiae are closely related to the base of Zone NP9. These species were grouped in the F. alanii group, in which the LO and HO can approximate the entry of Subzone NP9a and Subzone NP9b, respectively, when their zonal markers are absent or poorly preserved.
Furthermore, Blackites herculesii can approximate the top of Zone NP9. The LOs of D. mahmoudii, D. binodosus, D. diastypus, Zygrhablithus bijugatus, and Campylosphaera dela are not reliable bioevents in biostratigraphy. On the other hand, the LO and increased frequency of Neochiastozygus junctus can approximate the upper part of Zone NP9 and the lower part of Zone NP10.
Moreover, the LO of S. radians is a reliable bioevent for approximating the base of Zone NP11. Tribrachiatus orthostylus appears just below the base of Subzone NP10b. The abundance of calcareous nannofossils shows an increased frequency close to the P-E boundary. Ericsonia subpertusa dominates the calcareous nannofossil assemblages of the examined interval and suddenly increases above the P-E boundary. On contrast, a drop in the diversity of Fasciculithus is closely related to the P-E transition.
The P-E boundary at G. Nezzazat was placed at the base of Subzone NP9b that is denoted by the LOs of Discoaster araneus, D. anartios, and/or Rhomboaster spp. This boundary is coincident with the base of a black shale bed assigned to the DQM in the lower portion of the Esna Formation (Fig. 3). This reveals the presence of a small gap represented by the missing of DQB2 to DQB5. The variations in calcareous nannofossils reveal warming and oligotrophic conditions prevailed through the deposition of the upper Paleocene-lowermost Eocene at the study section.
Funding Open access funding provided by The Science, Technology & Innovation Funding Authority (STDF) in cooperation with The Egyptian Knowledge Bank (EKB).

Conflict of interest
The authors declare that they have no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.