The Maastrichtian–Thanetian strata exposed at the Dakhla Oasis is investigated to infer the inter–relationships between diversity, palaeooxygenation, palaeoproductivity, and palaeodepth. The proxies for diversity (Fisher’s α) (see Fig. 4), palaeooxygenation (ventilation) of the sea floor – BFOI (Benthic Foraminifera Oxygen Index) and % Oxiphilic taxa (see Fig. 4); for organic–flux to the sea floor (palaeoproductivity) include: percentage abundances of High organic–flux species (% HOFS; Fig. 4) and infaunal taxa; and benthic foraminiferal morphogroups (Figs. 5–6; Table 1) for reflecting the state of the benthic ecosystem. These approaches are then inferred in combination with the presence of characteristic benthic foraminiferal species and genera and paleodepth (Fig. 7).
[INSERT HERE – FIGURE 4]
One gram of sediment weight was used to pick benthic foraminifers that were soaked in the Na2CO3 solution before sieving them over 630, 125 and 63 μm mesh size, respectively. The 63–125 μm fraction size was used and identified under a binocular zoom stereomicroscope. A total of 315 samples (16,271 specimens) were analyzed. Of them, 195 were productive from the 315 m thick Maastrichtian–Thanetian succession exposed at the main escarpment of Gharb El–Mawohb (26° 01' 02" N, 28° 13' 18"E; Western Desert, Egypt) (see Farouk and Jain, 2016; Jain and Farouk, 2017).
However, the present study does have a one undeniable flaw that the specimen size (16,271 specimens from 315 samples, 195 being productive) is small. But, considering that there is still a large representation of diverse taxa (150 species), even in only 2 gram of sediment analyzed for this study, and the fact that a large number of samples were analyzed (315) (see Appendix–Tables 1–2), we believe that simple analysis and distribution patterns will reveal trends that will not be flawed. Owning to small specimen size, no rigorous analysis can be conducted such as Cluster Analysis, Factor Analysis and/or Principal Component Analysis, etc., as they require a minimum of 300 specimens per sample and as is the norm in most micropaleontological studies. Hence, here, a very basic correlation analysis (see Appendix–Tables 3–6) is conducted and vertical distribution trends are considered. The emphasis of the study, however, remains on assemblage analysis, corroborated by results of correlation analysis. This is also a reason why several proxies are considered for paleoenvironmental interpretation in addition to filed observations, P%, etc. Based on this, a tentative model is proposed for their inferred distribution (Fig. 8). The benthic foraminiferal changes at the Cretaceous/Paleogene (K/Pg) boundary are also analyzed (see Figs. 9–11, Tables 2-3; Appendix–Tables 7–8). Important benthic foraminiferal species are illustrated in Fig. 12 using scanning electron microscopy of the Geological Survey of Egypt.
[INSERT HERE – FIGURES 6-8, TABLE 1]
5.2. The Proxies
5.2.1. Diversity proxy
Fisher's α is used as a proxy for representing diversity (Fig. 4) as the latter and Shannon H are statistically significant (0.949, 0.000; correlation significant at the 0.01 level; 2–tailed). Fisher’s α, a within–habitat diversity index and a preferable proxy for species diversity, considers both evenness and richness. However, other species abundance models consider only evenness (Buzas and Gibson, 1969).
5.2.2. Oxygenation proxies
Two proxies are used to estimate the availability of oxygen within the sediment column (or to assess bottom waters as well as pore water oxygenation). These are Benthic Foraminiferal Oxygen Index (BFOI), and the percentage of Oxyphilic taxa (Fig. 4). It must be mentioned, in most continental slope and margin environments (as the present one), the depth of the oxic sediment layer rarely exceeds 5 cm and is mostly confined to the top cm of the sediment column (Corliss, 1985; Jorissen et al., 1998; Fontanier et al., 2002).
5.2.2.1. Benthic Foraminiferal Oxygen Index (BFOI)
Kaiho (1991), based on studies by Phleger and Soutar (1973), Bernhard (1986), Koutsoukos et al., (1990) and Perez–Cruz and Machain–Castillo (1990), developed an empirical ratio of oxic and dysoxic benthic foraminiferal morphotypes called the Benthic Foraminiferal Oxygen Index (BFOI). Based on benthic foraminiferal test morphology data from DSDP cores from the Pacific, South Atlantic and Indian oceans, Kaiho (1991) categorized them into three morphogroups, Oxic (O), Suboxic (S) and Dysoxic (D) (see Appendix–Table 1) and defined the index as [O/(O+D) x 100], where O is the number of oxic species and D the number of dysoxic species. When O = 0 and D+S >0 (S is the number of suboxic indicators), then the BFOI value is given by [(S/(S+D)–1] x 50 (Fig. 2). Later (Kaiho, 1994) calibrated modern values of the BFOI to levels of dissolved oxygen in bottom waters in the modern ocean (for details also Kaminski et al., 2002).
5.2.2.2. % Oxyphilic taxa
Jannink et al., (2001) proposed a transfer function (Oxygen content µMol/lt = 7.9602 + 5.95 x % oxyphilic taxa) to estimate the oxygen content within the sediment column (Fig. 4). Those taxa that occur in the topmost cm of the sediment column are considered oxiphylic, and their cumulative percentage is considered a proxy for bottom water oxygenation. The rationale being that, with increasing bottom water oxygenation, the availability of oxygen in the sediment also increases, resulting in an increased volume of the niche potentially occupied by oxiphylic taxa. In the present study, the epifaunal taxa (see Appendix–Table 1) are considered oxyphylic. The transitory epifaunal to shallow infaunal taxa are not included in the calculations as also species of Lenticulina and Osangularia. Lenticulina, an opportunistic epifaunal species, is an r–type strategist that thrives in slightly better oxygenation conditions with sufficient food availability, and displays a varied epifaunal to deep infaunal microhabitat preference (Tyszka, 1994; Jorissen et al., 1992; Reolid et al., 2008, 2012; Jurassic–Cretaceous). Whereas, representatives of the genus Osangularia (here: O. expanse and O. plummarae) have been noted to behaved opportunistically (Alegret et al., 2009). This is corroborated in the present study as they occur in Assemblage 8 (sample 152) and then suddenly appear just after the K/Pg event in Assemblage 11 (samples 203–211) and then disappear in the succeeding Assemblage 12 (sample 232) (Fig. 4).
Although, both BFOI and % Oxyphilic index employ % epifaunal taxa (= the Oxic (O) taxa used in BFOI; see Appendix–Table 1), however, BFOI also incorporates the presence of Suboxic (S) and Dysoxic (D) taxa not used in the % Oxyphilic index. Hence, BFOI considers a much broader community approach.
5.2.3. Palaeoproductivity proxies
Two palaeoproductivity proxies are used in this study – the percentages of benthic foraminiferal high organic–flux species (%HOFS) and infaunal taxa (Fig. 4).
5.2.1.1. High organic–flux species (% HOFS)
The following species are considered as high–organic–flux species in this study – Anomalinoides aegyptiacus, Bolivina cretosa, Bulimina prolixa, Nonionella africana, N. insect, Praebulimina kikapoensis, P. russi, Pyramidulina affinis, P. distans, P. semispinosa, P. vertebralis, P. zippei and Reussella aegyptiaca (see also Sen Gupta and Machain–Castillo, 1993; Sen Gupta, 1999; Fontanier et al., 2002; Gebhardt et al., 2004; Friedrich and Erbacher, 2006; Jorissen et al., 2007; Friedrich et al., 2009; Alegret and Thomas, 2013; Sprong et al., 2013) (see Fig. 4).
5.2.1.2. Percent infaunal taxa
The species are grouped under epifaunal and infaunal (shallow and deep infaunal with an undifferentiated group as infaunal) with a transitory group of epifaunal to shallow infaunal taxa (see Appendix–Table 1) (Fig. 4). The latter are not included in the infaunal group calculations. Data from various studies (Sliter, 1968; Sliter and Baker, 1972; Corliss, 1985; Jones and Charnock, 1985; Corliss and Chen, 1988; Bernhard, 1986; Langer, 1993; Rathburn and Corliss, 1994; Severin and Erskian, 1981; Jorissen, 1988; Linke, 1992; Sjoerdsma and Van der Zwaan, 1992; Jorissen et al., 1992; Linke and Lutze, 1993; Jorissen, 1988; Corliss and Emerson, 1990; Sjoerdsma and Van der Zwaan, 1992; Kaminski and Gradstein, 2005) have been used to construct Appendix–Table 2.
5.2.4. Morphogroups and subgroups
Regardless of taxonomy, the groupings of similar shapes or growth patterns of benthic foraminiferal tests reflect a particular type of environment, and hence, are good proxies for the prevailing palaeoenvironment, particularly for well–oxygenated temperate environments (Chamney, 1976; Jones and Charnock, 1985; Nagy, 1992; Kaminski et al., 1995; Nagy et al., 1995; Jones, 1999; Preece et al., 1999; van der Akker et al., 2000; Jones et al., 2005; Kender et al., 2008a, b; Cetean et al., 2011, Murray et al., 2011). In the present study (Figs. 6–7; Table 1), the benthic foraminiferal morphogroup assignment is after Koutsoukos and Hart (1990) (Table 1) as more recent schemes provide somewhat less differentiation (Nagy, 1992; Tyszka, 1994; Nagy et al., 1995; van der Akker et al., 2000; Reolid et al., 2008; Nagy et al., 2009; Cetean et al., 2011; Chan et al., 2017). Most of these aforementioned studies are either from the Jurassics (Nagy, 1992; Tyszka, 1994; Nagy et al., 1995; Nagy et al., 2009) or from a different time period (Cetean et al., 2011, Santonian–Campanian; Chan et al., 2017, Tertiary) and thus, have fewer common species to correctly interpret morphotypes. Those that are from a similar period (van der Akker et al., 2000, Campanian–Maastrichtian) are largely from bathyal settings, contrary to the neritic one in the present study and hence, have fewer common species for comparison. The closest comparative account is either that of Tyszka (1994) or of Koutsoukos and Hart (1990). The latter have the maximum number of common species with detailed explanation of their habitat preferences. Thus, based on this latter dataset, the species from the present study are interpreted for their microhabitat assignment. The differentiation of morphological categories (Table 1) is based on three criteria: (a) test morphology, comprising general outline, number of chambers, chamber arrangement and architecture; (b) life habitat, characterized as epifaunal, shallow infaunal and deep infaunal (few are transitional – epifaunal to shallow infaunal); and (c) their supposed feeding strategies, characterized as suspension feeders, deposit feeders, herbivores or bacterial scavengers.
5.2.5. Wall structure (proxy for Palaeodepth)
To infer bathymetry, the agglutinated foraminiferal dataset is grouped based on their preference for palaeodepth i.e. into simple–walled arenaceous agglutinated foraminifera assemblage (representing littoral environment with fresh water supply), complex–walled arenaceous agglutinated foraminifera assemblage (representing deeper littoral environment with normal marine water), and calcareous agglutinated foraminifera assemblage (representing shelf environment) (see also Berggren, 1974; Luger, 1985, 1988; Cherif and Hewaidy, 1986; Orabi, 1995, 2000) (Fig. 7). This is further corroborated with bathymetry inferred from %Planktic foraminifera (%P), total calcareous benthic foraminifera and characteristic species (Fig. 7).
All the above mentioned proxies are subjected to basic statistical analysis (Pearson correlation) (Appendices Tables 3-6) and evaluated based on the framework of the 19 benthic foraminiferal assemblages (species abundance pattern) as established by Farouk and Jain (2016, 2017) and Jain and Farouk (2017) (see Fig. 7).