Paleoenvironmental changes across the Paleocene–Eocene boundary in West Central Sinai, Egypt: geochemical proxies

A geochemical analysis has been conducted on twenty-six sediment samples spanning the P–E boundary interval collected from the Esna Shale in three well-dated stratigraphic sections in west-central Sinai, Egypt to interpret paleoenvironmental changes associated with the P–E boundary events. The Esna Shale consists of hemipelagic marine shales and marls and it is subdivided stratigraphically into the uppermost Paleocene Hanadi Member unconformably overlain by the lowermost Eocene Dababiya Quarry Bed (DQB) and El-Mahmiya members. A variety of geochemical proxies including the Al-normalized elemental concentrations and redox-related elemental ratios and parameter (V/Cr, V/V + Ni and Mn*) and productivity indicators (Porg and Babio) were employed for paleoceanographic interpretations. Across the P–E transition, the concentrations of SiO2, TiO2, Al2O3 Fe2O3, MgO, V, Cr, Ni, Cu, Rb, and Zr notably increased, whereas the concentrations of CaO, MnO, Sr, and Zn abruptly decreased. The variation in the elemental concentrations is attributed to the carbonate dissolution because of increased ocean acidification as well as a brief increase in the detrital influx associated with the brief humid interval at the base of the Eocene. The Al-normalized detrital-related elements (Ti, Zr, and Rb) values show a relatively homogeneous profile suggesting a uniform detrital input from an unchanged source rock. The simultaneous significant increase in the V/Cr and V/V + Ni ratios and enrichment of Ni, Cu, and Zn in the sediments of the DQB suggest that the depositional basin experienced dysoxic to slightly anoxic bottom conditions. The coeval increase in the Porg and Babio in the sediments of the DQB suggests an increase in nutrients availability and consequently primary productivity possibly due to enhanced upwelling during early Eocene.


Introduction
The Paleocene-Eocene boundary (P-E) interval (56 Ma ago) known as the Paleocene-Eocene thermal maximum (PETM; Kennett and Stott 1991) is a significant and short-lived (170 to ~ 217 kyr; Röhl et al. 2007;Murphy et al. 2010) extreme and abrupt global warming event marked by an abrupt increase in the sea surface and deep-sea temperatures by 5-8 °C and 6 °C, respectively (Sluijs et al. 2006;Zachos et al. 2006). This event had a profound impact on the atmospheric and oceanic circulation patterns and marine and terrestrial ecosystem. The sedimentary records across the PETM bear several lithological, mineralogical, geochemical and biological criteria explaining the environmental perturbations and ecosystem readjustments associated with this dramatic climatic warming (Gawenda et al. 1999). The most prominent criteria of this extreme global event are: (1) a remarkable negative (− 2 to − 7‰) carbon isotope excursion (CIE; Kennett and Stott 1991;Kender et al. 2012), (2) a rise of water salinity, (3) a drop of pH and dissolved oxygen and I. M. Ghandour shoaling of the CCD (Sluijs et al. 2007;Zachos et al. 2008;Westerhold et al. 2011;Keller et al. 2018), (4) extinction of 35-50% of deep-sea benthic foraminifera (Alegret et al. 2009), (5) turnover of planktonic foraminifera and calcareous nannofossils (Aubry et al. 2007;Self-Trail et al. 2012;Cao et al. 2018) and (6) diversification and migration of terrestrial mammals (Bowen et al. 2002). In addition, the high abundance of kaolinite and abrupt increase in eutrophic dinoflagellates in many shallow marine sedimentary sections spanning the PETM (Bolle andAdatte 2001, Crouch et al. 2003) suggest increased precipitation and runoff and enhanced marine productivity. The mechanism(s) and consequences of the PETM global changes has been a focus in several research efforts since its discovery.
Several terrestrial and marine sedimentary records spanning the PETM with characteristic carbon isotope excursion (CIE) and distinctive lithologic and biotic features have been documented at several localities worldwide (Speijer and Wagner 2002;McInerney and Wing 2011;Pälike et al. 2014;Schoon et al. 2015;Cao et al. 2018). Among of them, the Tethys is still the key region for investigating the mechanism of P-E global change (Gawenda et al. 1999;Soliman et al. 2011;Khozyem et al. 2013Khozyem et al. , 2015Stassen et al. 2012;Gavrilova et al. 2018). The most extensive and continuous sedimentary record across the P-E boundary is documented from the southern margin of the Tethys. This record is known in Egypt as the Dababiya Quarry Beds (DQB) that was first described from the Dababiya, Luxor. In 2003, the DQB was selected as the Global Boundary Stratotype Section and Point (GSSP) for the basal Eocene (Dupuis et al. 2003). The DQB display distinctive lithological, biological, mineralogical and chemical attributes. With little variations from the type section, the DQB is recognized from the Paleogene sections in west-central Sinai (Khozyem et al. 2013;Faris et al. 2015;Obaidalla et al. 2017). Studies carried out in west-central Sinai concerned mainly with foraminiferal (Obaidalla et al. 2017) and calcareous nannoplankton (Abu Shama and Faris 2007;Faris et al. 2015) biostratigraphy. Paleoenvironmental, mineralogical and geochemical studies on the PETM in west-central Sinai are relatively rare (Bolle et al. 2000;Speijer and Wagner 2002;Khozyem et al. 2013).
The present study utilizes inorganic chemical composition of the P-E sedimentary record in west-central Sinai, Egypt to (1) reconstruct paleoenvironmental changes associated with the P-E boundary global perturbations in southern Tethys, (2) interpret elemental enrichments in terms of environmental response to the P-E boundary global events and (3) correlate the findings with the previous investigations in other Tethyan margin sections.

Geological background
At the end of the Paleocene, the Tethyan Ocean was an extensive northward deepening epicontinental basin covering most of Egypt, Israel, and Jordan. The southern margin of the Tethys was located in the northern tropical zone that was strongly affected by intermittent upwelling episodes (Speijer and Wagner 2002;Schulte et al. 2011). Many well preserved and widely distributed stratigraphic sections in North Africa and Arabian Craton contain biotic and geochemical evidence related with these upwelling episodes (Almogi-Labin et al. 1993). These sections display similar and distinctive features revealing the P-E major global changes. The most extensive and continuous P-E sedimentary records of distinctive lithologic, biotic and chemical characteristics at the southern margin of the Tethyan seaway are widely distributed in Egypt. In west-central Sinai, these strata are represented by continuous succession of hemipelagic sediments containing abundant and diverse microfossil assemblages (Bolle et al. 2000;Speijer and Wagner 2002;Khozyem et al. 2013;Faris et al. 2015;Obaidalla et al. 2017).
In west-central Sinai, the P-E upper bathyal (500-600 m water depth) grey to dark grey shales belonging to the lower Paleogene Esna Shale are well exposed at Wadi (W.) Matulla, W. Feiran and Gebel (G.) Mekattub (Fig. 1). The stratigraphy of these deposits is well constrained subdividing the Esna Shale into three members; the uppermost Paleocene Hanadi and the lowermost Eocene Dababiya Quarry Bed (DQB) and El-Mahmiya. The sediments of the Hanadi Member are dominated by light brown to light grey massive mudstones, varying in thickness laterally from 1.8 to 6 m because of syn-depositional faulting (Faris et al. 2015). The Dababiya Quarry Bed (DQB) Member is a sharp-based distinctive horizon varying in thickness from 0.5 to 0.9 m. It consists at the base of dark grey massive shales (3-12 cm thick) sharply overlain by light to medium grey silty mudstone (~ 25 cm thick) with abundant coprolites, phosphate nodules and fish remains. The last bed contains Chondrites trace fossils (Faris et al. 2015). The upper part of the DQB consists of marl bed up to 40 cm thick. The P-E boundary is placed at the base of a dark grey shales that define the contact between the Hanadi and Dababiya Quarry members (Bolle et al. 2000;Speijer and Wagner 2002;Khozyem et al. 2013). The dark grey clay layer is known at W. Nukhul section by Khozyem et al. (2013) as condensed clay-rich interval. The base of dark grey shales correlates with the base of the carbon isotope excursion (CIE) level. This level is consistent with the base of the Eocene Series in the Dababiya (GSSP) in Egypt (Dupuis et al. 2003;Aubry et al. 2007). Calcareous nannofossil biostratigraphy (Faris et al. 2015) suggests a short hiatus (absence of the NP9b subzone) at the P-E boundary. Paleoenvironmental changes across the P-E boundary, West Central Sinai, Egypt The DQB rests unconformably over the latest Paleocene Hanadi Member with no evidence for subaerial exposure (Schulte et al. 2011). The sediments of the Hanadi Member were deposited during the late Eocene sea-level high stand (HST), whereas the sediments of the DQB were deposited during the early Eocene rapid sea-level rise (TST) with poorly preserved lowstand deposits (Speijer and Wagner 2002).
The paleoclimate of the P-E boundary interval in west-central Sinai was interpreted relying on clay minerals analysis (Bolle et al. 2000;Khozyem et al. 2013). The enrichment of palygorskite, sepiolite, and illite in the sediments of pre-PETM at W. Nukhul west-central Sinai (Khozyem et al. 2013) suggests a dominantly arid climate, whereas the abrupt increase of the kaolinite content in the lowermost sediments of the PETM indicates a short period of warm humid conditions. The increase in the smectite content in the lower Eocene sediments suggests dominantly arid climate punctuated by short intensive humid intervals (Khozyem et al. 2013).

Materials and methods
Twenty-six fresh rock samples spanning the P-E boundary interval were collected from W. Matulla (9 samples), G. Mekattub (9 samples) and W. Feiran (8 samples) westcentral Sinai (Fig. 1). Conventional XRF technique was conducted at the Department of Geosciences, Osaka City University, Japan to determine the chemical composition of the sediments. Glass beads were prepared by fusing 1.8 g of dry powdered sample with 3.6 g of spectroflux (Li 2 B 4 O 7 20%, LiBO 2 80%), 0.54 g of oxidant LiNO 3 , traces of LiI and 4.2 g of Li 2 B 4 O 7 at 800 °C for 120 s and 1200 °C for 200 s (Tawfik et al. 2017). The analysis was carried out using a RIGAKU RIX 2100 X-ray fluorescence spectrometer (XRF), equipped with Rh/W dualanode X-ray tube under 50 kV and 50 mA accelerating voltage and tube current, respectively. The accuracy of the analyses was estimated to be ± 2-3% for major elements and ± 10-15% for trace elements. Precision was measured at < 0.5% for major elements and < 7% for trace elements. Total iron is introduced as Fe 2 O 3 t and major and trace elements are reported as wt% and µg/g, respectively. Trace-element concentrations are normalized to Al to correct dilution by organic matter and authigenic minerals, to compensate the effect of mineralogy and grain size on elemental contents and to differentiate the detrital, biogenic and authigenic origin of the elements (e.g., Riquier et al. 2006;Soua et al. 2011;Khozyem et al. 2015). The enrichment factor (EF) of the environmentally significant elements is determined (Algeo and Maynard 2004;Rimmer 2004) by normalizing the element against the average shale composition (Li and Schoonmaker 2003). The palaeoenvironmental conditions during the P-E boundary interval is evaluated using elemental ratios indicative of detrital input (Ti/Al, Rb/Al and Zr/Al), productivity-related elements and ratios (P org and Ba bio , Ni/ Al, Cu/Al and Zn/Al), redox-sensitive elemental proxies (Mn*, V/Cr and V/V + Ni).

Results
The range and mean values of major oxides and trace elements of the sediments of the P-E boundary interval in west-central Sinai, Egypt are shown in Table 1. The chemical composition of the DQB differs remarkably from the composition of the underlying and overlying Hanadi and El-Mahmiya members, respectively.

Enrichment factors (EF)
The degree of enrichment of the environmentally sensitive major and trace elements (Mn, P, V, Cr, Ni, Cu, Zn and Ba) is determined using enrichment factor (EF; Calvert and Pedersen 1993;Arnaboldi and Meyers 2003;Rimmer 2004;Soliman et al. 2011). It is estimated by normalizing the concentration of the element (X) relative to Al in the sediments and comparing these ratios to the same ratio in average shale. The enrichment factor is determined using the average shale of Li and Schoonmaker (2003) as follows: The major and trace elements of the studied sediments show different levels of enrichment. Of particular note are the wide variation of the EF for V, Cr, Ni, Cu, and Zn.
The sediments of the DQB exhibit a wide range of EF values, but some samples are highly enriched in V, Cr, Ni, Cu and Zn (2-124 fold). At W. Matulla, the ranges of the EF values of V, Cr, Ni, Cu and Zn in the sediments of the EF = X/Al sample /X/Al shale .
DQB are 2-116, 3-53, 8-91, 7-21 and 3-53 ( Fig. 3 and Table 3). The enrichment factor values of V, Cr, Ni, Cu and Zn range in the sediments of the DQB at G. Mekattub section from 4 to 124, 3 to 41, 11 to 123, 2 to 15 and 2 to 56, respectively and at W. Feiran section vary from 3 to 15, 2 to 16, 8 to 45, 1 to 10 and 1 to 28, respectively ( Fig. 3 and Table 3). The values of the enrichment factor for Mn are slightly low but highly variable varying from 0.05 to 23. The lowest EF values for Mn (0.05, 0.16 and 0.13) were obtained from the dark grey shale layer at the base of the DQB at W. Feiran, G. Mekattub and W. Matulla sections, respectively (Fig. 3).
The enrichment factor values of Ba and P varied remarkably in the studied sections. The highest enrichment factor values of Ba were recorded generally in the sediments of the DQB at W. Feiran (162 and 170) and W. Matulla (177) sections, whereas the sediments of the DQB at G. Mekattub section are highly depleted in Ba (EF < 1). The sediments of the DQB display moderate to high EF values of P at W. Matulla (6-140), G. Mekattub (11-93) and W. Feiran (5-31). Similarly, the EF values for P in the sediments of the Hanadi Member vary from 9 to 10, 16 to 19 and 14 to 15 at W. Matulla, G. Mekattub and W. Feiran sections, respectively, whereas they vary in the sediments of El-Mahmiya Member from 15 to 22, 4 to 7 and 5 to 7, respectively (Table 3).
Generally, the EF P values are relatively high in the sediments of the DQB comparing to the underlying and overlying sediments. However, the dark grey shales at the base of the DQB are highly depleted in P ( Fig. 3 and Table 3). The EF P values are higher at W. Matulla and G. Mekattub sections than at W. Feiran section (Fig. 3). The Ba is highly depleted in the sediments of the P-E boundary. However, a single peak (EF Ba = 177) was recorded from the DQB sediments at W. Matulla section and two successive samples from the DQB at W. Feiran section show EF Ba of 162 and 170 (Table 3).

Redox proxies
The redox-sensitive trace elements (e.g. V, Cr, Ni, Cu, and Zn) behave differently under different redox conditions and they are enriched under anoxic (oxygenfree) sediment pore and bottom water conditions. The enrichment of the redox-sensitive trace elements (V, Cr, Ni, Cu, and Zn) and their ratios (V/Cr and V/V + Ni) in marine sediments can be utilized to infer paleoredox conditions (Algeo and Maynard 2004;Tribovillard et al. 2006;Fu et al. 2014). Values of V/Cr ratios < 2 indicate oxic conditions, from 2 to 4.25 indicate dysoxic and > 4.25 show suboxic to anoxic. On the other hand, values of V/(V + Ni) ratio < 0.46 indicate oxic conditions, 0.46-0.60 dysoxic and greater than 0.6 anoxic. The V/Cr ratio is the sediments of the Hanadi and El-Mahmiya members are generally < 2 and the V/V + Ni ratio are generally ≤ 0.5. The highest values of the V/Cr and V/V + Ni ratios were obtained from the same level in the sediments of the DQB (Table 3, Fig. 4). The V/Cr ratio varies in the sediments of the DQB from 1.29 to 2.76, from 1.38 to 4.41 and from 1.07 to 3.2 at W. Feiran, G. Mekattub and W. Matulla sections, respectively. Similarly, the range of the V/V + Ni ratios in the DQB at W. Feiran, G. Mekattub and W. Matulla sections are 0.46-0.65, 0.44-0.77 and 0.4-0.81, respectively (Table 3).
Another parameter that has been used as redox indicator is the Mn* (Cullers 2002). It is calculated as follows: Mn * = log Mn sample /Mn shale /Fe sample /Fe shale ] The negative values indicate sub-oxic to anoxic bottom conditions, whereas positive Mn* values indicate deposition under oxic and/or nearly suboxic conditions (Cullers 2002). The Mn* value varies in the three sections respectively from 0.81 to 0.93, 1.03 to 1.35 and 1.12 to 1.58. The minimum values (0.16, 0.26 and 0.25) were recorded from the dark grey shale at the base of the Eocene strata (the dissolution interval in the description of Khozyem et al. 2013) (Table 3, Fig. 4). Under low oxygen depositional conditions, Mn is depleted, whereas values close to average shale or slightly above are common deposited under oxic conditions (Brumsack 2006

Productivity proxies
A variety of geochemical proxies have been utilized to reconstruct paleoproductivity, among of them, the organically derived phosphorus (P org ) and biogenic barium (Ba bio ) are the most widely used (Brumsack 2006;Algeo and Ingall 2007;Schoepfer et al. 2015). Inferences concerning paleoproductivity are commonly based on variations in the fluxes rather than absolute/ total concentration data (e.g., Algeo et al. 2011;Schoepfer et al. 2015 and references therein). Under conditions favoring their preservation, the fluxes of P org , and Ba bio (or authigenic barite) may serve mutually or individually as proxies for productivity (Paytan and Griffith 2007;Schoepfer et al. 2015). The organic phosphorus fluxes (P org ) rather than total phosphorus (P or P total ) is estimated as follows: The detrital (P/Al) det ratio of 0.0087 is assumed based on the average P and Al concentrations of the upper continental crust (McLennan 2001).
The values of P org varied in the sediments of the DQB from 0.38 to 2.61%, 0.75 to 2.48 and 0.31 to 1.5% at W. Matulla, G. Mekattub and W. Feiran sections, respectively (Table 3). The values of P org in the sediments of the Hanadi and El-Mahmiya vary generally from 0.17 to 0.73% (Table 3, Fig. 4). The high values of EF P in the sediments of the DQB are consistent with the enrichment of these strata with phosphatic nodules and fish remains (Faris et al. 2015).
On the other hands, the biogenic Ba (Ba bio ) that includes all non-detrital Ba was determined as follows: The (Ba/Al) det ratio of crustal composition that ranges from 0.005 to 0.010 (e.g., Taylor and McLennan 1985) is used. Negative Ba bio values may suggest that the Ba in the sediments are generally of detrital origin. The values of biogenic Ba (Ba bio ) are generally high in the sediments of the W. Feiran section. It covers a wide range 4115-24,759 µg/g, whereas the majority of samples from G. Mekattub and W. Matulla sections are represented by non-biogenic Ba (Table 3). The sediments of the Hanadi and the basal part of the DQB display Ba bio concentrations ranging from 2940 to 5434 µg/g and a single peak of Ba bio (27,430 µg/g) was recorded from the DQB at W. Matulla section (Fig. 4). The wide variability of Ba bio content indicates that a considerable part of the Ba tot is composed of biogenic barium (Ba bio ), confirming that some intervals experienced high productivity (Table 3).

Fig. 4
Vertical variation in the redox proxies (V/Cr, V/V + Ni and Mn*) and the concentrations of productivity proxies (P org . and B abio ). The vertical variations in V/Cr and V/V + Ni reveal that the DQB sediments were deposited under dysoxic to anoxic bottom conditions. The positive excursion in the P org immediately above the dark grey shales demarcating the P-E boundary suggests increased primary productivity. The lithostratigraphic units and calcareous nannofossil biozones are shown in Fig. 1 The Al-normalized Cu, Ni, and Zn, which are strongly related to organic matter, show low values in the sediments of the Hanadi and El-Mahmiya members comparing to the sediments of the DQB (Table 2, Fig. 2). Below the P-E boundary, the average values of Al-normalized Ni, Cu and Zn are 41, 4 and 55 × 10 −4 at W. Feiran section, 46, 3 and 46 × 10 −4 at G. Mekattub and 43, 7 and 62 × 10 −4 at W. Matulla section. Immediately above the P-E boundary, the Ni/Al, Cu/Al, and Zn/Al values varied dramatically among the studied sections. Cu/Al values suddenly increased at the studied section with the highest values recorded from W. Feiran. Across the P-E boundary Ni/Al and Zn/Al increased abruptly at W.  (Table 2).

Discussion
In the present study, the geochemical data of the P-E hemipelagic marine sediments in west-central Sinai is of special interest to interpret the environmental response to the PETM global changes in the southern Tethyan region. The geochemical characterization of the P-E sediments shows that the sediments comprise mainly of admixture of detrital siliciclastic and calcareous biogenous components. In a context of relatively uniform and small variability in the terrigenous related elements, the uniform and relatively limited range of the Al-normalized Si, Ti, Zr, Rb and Nb values suggest that the source rock composition remained unchanged during the P-E boundary interval. The dominantly homogenous finegrained sediments minimize the effect of hydraulic sorting on sediment composition. Therefore, the elemental enrichment/depletion across the P-E boundary is possibly related to changes in environmental parameters, particularly bioproductivity and bottom redox conditions. The abrupt increase in the concentrations of major oxides (SiO 2 , TiO 2, and Al 2 O 3 ) and the coeval depletion in the CaO concentration across the P-E transition are attributed to the CaCO 3 dissolution due to increased oceanic acidity (Zachos et al. 2005).
In sediments deposited under such extreme conditions, redox-sensitive trace elements may be strongly enriched (by factors of > 100) in comparison to the average marine shale (März et al. 2008). The Al-normalized redox-sensitive elements (V, Cr, Ni, Cu, and Zn), the ratios (V/Cr and V/V + Ni) and the Mn* values display a relatively similar profile with the absence of significant oscillation throughout the Hanadi and El-Mahmiya members. An interval in the DQB show an extreme elemental enrichment (particularly for V, Ni, Cu and Zn), reflecting a brief period of fully dysoxic to fully anoxic bottom conditions (Algeo and Maynard 2004). Finally, elemental enrichment factors drop to or near the background values within El-Mahmiya Member sediments, suggesting that normal oxygenation like before the PETM was restored. These proxies suggest dominantly oxic to slightly dysoxic bottom conditions. In contrast, these proxies are highly fluctuated in the sediments of the DQB from dysoxic to anoxic and rarely oxic conditions. There is a slight contradiction among these proxies (Figs. 4 and 5). This contradiction is recorded by Khozyem et al. (2013;Fig. 7). The lowest Mn* and EF Mn values were recorded from the dark grey shales at the base of the DQB. The low values are attributed to the removal and diffusion of Mn from sediments to water column under strongly oxygen-depleted bottom waters (Bruland 1983;Cullers 2002;Tribovillard et al. 2006). However, this suggestion contradicts with the corresponding low values of V/Cr and V/V + Ni ratios that show dominantly oxic bottom conditions. The rest of samples show relatively high Mn* values indicating deposition under conditions that are more oxic and this assumption contradicts the dysoxic to anoxic inference by the relatively high values of V/Cr and V/V + Ni ratios. The dark grey shales at the base of the DQB (the condensed clay interval; Khozyem et al. 2013) correspond to a sudden drop in the CaO concentration, possibly correlated with the CaCO 3 dissolution interval (Khozyem et al. 2013). The positive correlation between MnO and CaO (r = 0.8) suggests that the Mn resides in the carbonate phase, most likely as Mn coatings (Thurow et al. 1992). Therefore, the distribution of Mn in the P-E sediments is primarily controlled by the carbonate content rather than redox conditions. Under oxic-dysoxic conditions, Ni and Cu are deposited mainly in association with the detrital fraction, whereas under anoxic conditions, they are deposited mainly in the form of organometallic complexes. . Enrichments of redoxsensitive trace elements (V, Cr, Ni, Cu, and Zn) compared to the average shale level within the DQB indicate that pore-and bottom-water was occasionally sub-oxic to anoxic.
Geochemical and micropaleontological data of several sections spanning the PETM in the Peri-Tethys regions characterizing anoxic bottom conditions were recognized from Crimea and Uzbekistan (Gavrilova et al. 2018) and Egypt (Speijer et al. 1997;Speijer and Wagner 2002;Soliman 2003;Soliman et al. 2011;Schulte et al. 2011;Khozyem et al. 2013Khozyem et al. , 2015. A single anoxic event spanning the PETM was recorded at G. Qrieya (Soliman 2003), Dababiya GSSP (Schulte et al. 2011) and W. Nukhul (Khozyem et al. 2013). Khozyem et al. (2015) recognized two anoxic events demarcated by negative Ce-anomalies at Dababiya GSSP and associated with enrichment of redox-sensitive elements and the formation of idiomorphic and framboidal pyrite. The development of bottom anoxicity during the PETM was attributed to: (1) the inflow of poorly oxygenated intermediate water into the epicontinental circulation combined with enhanced upwelling and increased primary productivity (Speijer and Wagner 2002), (2) water-column salinity/temperature stratification and eutrophication (Schulte et al. 2011;Khozyem et al. 2015) and (3) increased ocean acidification due to methane hydrate release (Khozyem et al. 2015). Water column stratification prevented light penetration and water ventilation and increased surface water productivity and organic matter influx to the seafloor, which in turn led to consumption of bottom oxygen by bacteria (Schulte et al. 2011;Khozyem et al. 2015). Fluctuations between anoxic and suboxic or dysoxic bottom waters were inferred by trace fossils (Faris et al. 2015) and by the benthic foraminiferal records in the DQB sediments (Speijer and Wagner 2002).
The contents and the vertical trends of P org and Ba bio across the P-E boundary vary among the studied sections and show inconsistent relationship suggesting a local influence on the behaviour of the two elements. The occasional high positive excursion of Ba bio (up to 2.7%) in the sediments of the DQB at W. Feiran and W. Matulla sections compared with the concentration in average shale (0.058%) possibly indicate increased organic productivity during the earliest Eocene in west-central Sinai. This scenario is supported by increased nutrient supply through coeval enrichment of P org (up to 2.6%). The relatively high P org content in the basal sediments of the DQB coincides with the low Mn* value suggesting that bottom condition was not anoxic (Fig. 4). Phosphate (P) is generally depleted under anoxic condition because it is easily remobilized from the sediments and diffuses back to water column. The negative correlation with detritalrelated elements (Ti and Zr) suggests that the coeval enrichment of the Ni, Cu, and Zn in the sediments of the DQB further supports enhanced primary productivity during earliest Eocene through the formation of organometallic complexation with organic matter and/or the removal of surface waters by plankton growth. Several micropaleontological and geochemical proxies at several oceanic sites support the idea of increased upwelling and fluxes of nutrients enhanced primary productivity during the PETM (Schmitz et al. 1997a, b;Bains et al. 2000;Crouch et al. 2003). The abundance of tiny biogenic barite crystals (up to tens of microns) and opal-CT within the P-E boundary sediments at W. Matulla and Abu-Zenima suggests enhanced primary productivity (Bolle et al. 2000). At the Dababiya (GSSP), the occurrence of pyrite, barite, phosphate concretions, high TOC combined with enrichment of the productivity-related elements (P, Zn, and V) in the sediments of the PETM suggest increased primary productivity (Soliman et al. 2011;Khozyem et al. 2015). Enhanced productivity was attributed to an increase in the input of detrital minerals and organic matter (i.e., fossilized wood; Dupuis et al. 2003) to the shallow zones of the southern Tethyan margin basin. The complex mixture of clay minerals recorded at the Dababiya (Bolle et al. 2000;Schulte et al. 2011;Soliman et al. 2011;Khozyem et al. 2015) suggests that intensive chemical weathering and erosion of sources on the continent by the late Paleocene-Early Eocene sea-level and climate changes (Dupuis et al. 2003). The Early Eocene climate models suggested that the Tethyan southern margin has been affected by intermittent upwelling episodes initiated by increased atmospheric contrast between the arid, sub-tropical coastal zone and the humid, tropical continent intensifying the off-shore directed Ekman transportation (Speijer and Wagner 2002). The presence of condensed phosphate in the middle part of the DQB is consistent with upwelling activity (Bolle et al. 2000).
Bottom-water redox condition exerts a strong influence on the accumulation of productivity-sensitive elements and there is interplay between productivity and bottom redox conditions . Under reducing conditions, the burial fluxes of P org and Ba bio are diminished owing to the reductive dissolution of Mn-Fe-oxyhydroxide particles, thus releasing adsorbed P, and barite crystals, releasing Ba (Tyson 2005), whereas oxidizing conditions facilitate the preservation of P org and Ba bio within the sediments. At G. Mekattub (sample Mk-5), a coeval positive excursion in P org , EF P , V/Cr, and V/V + Ni. Similarly, at W. Matulla section, sample Mt-5 displays high values of EF P and EF Ba , P org and Ba bio content corresponding to relatively high values of V/Cr and V/V + Ni. The dysoxic to slightly anoxic bottom conditions were not severe enough to fully release the P and Ba from sediments. Some of the released P and Ba are possibly retained in the sediment and ultimately incorporated into authigenic P-and Ba-bearing mineral phases. High P accumulation rates during the Cenomanian-Turonian oceanic anoxic event (OAE2) was associated with high organic influx suggesting that P accumulation is a good indicator for productivity changes unless the bottom conditions change to severely anoxic (Mort et al. 2007).

Conclusions
Geochemical analysis of the Esna Shale sediments at three well-dated stratigraphic sections in the west-central Sinai, Egypt enabled interpreting environmental changes across the Paleocene-Eocene transition. Samples were collected from the uppermost Eocene Hanadi Member and the lowermost Eocene Dababiya Quarry Bed (DQB) El-Mahmiya members. Geochemical proxy investigation of the P-E boundary interval revealed the following conclusions: 1. The concentrations of major (SiO 2 , TiO 2 , Al 2 O 3 Fe 2 O 3 , MgO) and trace (V, Cr, Ni, Cu, Rb, and Zr) elements display remarkable positive excursion, whereas the concentrations of CaO, MnO, Sr, and Zn show abrupt negative excursion. 2. The Al-normalized environmental-sensitive elements (Mn, P, V, Cr, Ni, Cu, Zn, and Ba) exhibit a relatively uniform trend in the sediments of the Hanadi and El-Mahmiya members. On the other hand, they are highly enriched in the sediments of the DQB. 3. At the very base of the PETM event beds (transition from Hanadi to DQB), the dark grey shale is virtually depleted in CaO suggesting a brief period of carbonate dissolution, possibly associated with global ocean acidification. 4. The enrichment of the redox-sensitive trace elements and the their ratios (V/Cr and V/V + Ni) within the sediments of the P-E boundary interval suggest normal oxygenated conditions during deposition of the El-Hanadi Member followed by a gradual onset of oxygen depletion during the deposition of DQB. The sediments of the DQB show high overall V/(V + Ni) values and data plot within the suboxic and rarely anoxic fields, whereas the Hanadi and El-Mahmiya samples consistently show lower values than do the DQB. Similarly, V/Cr values show similar differences, and values are weighted towards suboxic to anoxic conditions for the DQB and towards oxic conditions for the Hanadi and El-Mahmiya. 5. A significant enrichment in P org and Ba bio contents in the sediments of the DQB indicates increased primary productivity induced by increasing nutrients supply through terrestrial runoff and enhanced upwelling process. 6. The enrichment in productivity-related elements in line with enriched redox-sensitive elements suggests that the reducing bottom conditions were not severe enough to fully release and diffuse back the P and Ba from sediments to seawaters.