New Biochronological Scales of Planktic Foraminifera for the Early Danian Based on High-Resolution Biostratigraphy

After the Cretaceous/Paleogene boundary (KPB) catastrophic mass extinction event, an explosive evolutionary radiation of planktic foraminifera took place in consequence of the prompt occupation of empty niches. The rapid evolution of new species makes it possible to establish high-resolution biozonations in the lower Danian. We propose two biostratigraphic scales for low-to-middle latitudes spanning the first two million years of the Danian. The first is based on qualitative data and includes four biozones: the Guembelitria cretacea Zone (Dan1), the Parvularugoglobigerina longiapertura Zone (Dan2), the Parvularugoglobigerina eugubina Zone (Dan3), and the Parasubbotina pseudobulloides Zone (Dan4). The latter two are divided into several sub-biozones: the Parvularugoglobigerina sabina Subzone (Dan3a) and the Eoglobigerina simplicissima Subzone (Dan3b) for the Pv. eugubina Zone, and the Praemurica taurica Subzone (Dan4a), the Subbotina triloculinoides Subzone (Dan4b), and the Globanomalina compressa Subzone (Dan4c) for the P. pseudobulloides Zone. The second scale is based on quantitative data and includes three acme-zones (abundance zones): the Guembelitria Acme-zone (DanAZ1), the Parvularugoglobigerina-Palaeoglobigerina Acme-zone (DanAZ2), and the Woodringina-Chiloguembelina Acme-zone (DanAZ3). Both biozonations are based on high-resolution samplings of the most continuous sections of the lower Danian worldwide and have been calibrated with recent magnetochronological and astrochronological dating.


Introduction
The high biostratigraphic resolution of planktic foraminifera is a result of their rapid evolution. The first and last appearances of planktic foraminiferal species occur within a relatively short period of geological time, usually providing us with detailed biochronological records [1,2]. A large number of biostratigraphic horizons (biohorizons) can be recognized, but not all of them are of any utility in biochronostratigraphy and biochronology. Only a few species, the so-called index species, allow robust biochronostratigraphic correlation, due to their easy taxonomic distinction, high abundance, short chronological and stratigraphic distribution, wide biogeographic distribution, and high preservation potential. The lowest and highest occurrence data (LOD and HOD) of index species are the most widely used key-biohorizons for defining planktic foraminiferal biozones and subbiozones, these being mainly of two kinds: range zones (with two subtypes: taxon-range and concurrent-range zones) and interval zones (with three subtypes: lowest-occurrence, highest-occurrence, and partial-range zones). Other kinds of biozone, such as abundance zones (acme-zones), are not commonly used for planktic foraminiferal zonations.
The highest resolution of the planktic foraminiferal biostratigraphic scales is achieved in the lower Danian, after the Cretaceous/Paleogene boundary (KPB) mass extinction, because numerous small-size species began to appear following a model of "explosive" evolutionary radiation [3,4]. Planktic foraminiferal specialists have tended to establish

Reference Sections, Key Biohorizons, and Calibration Methods
The two lower Danian biozonations proposed here are mainly based on the pelagic sections of Caravaca [8,24,[30][31][32], Zumaia [9,24,30,33] and Agost [24,34] (Spain), Gubbio [9,35,36] (Italy), El Kef [24,[37][38][39][40], Aïn Settara [24,30,39,41,42] and Elles [24,30,43] (Tunisia), La Lajilla [24,44] and Bochil [30,45] (Mexico), and Moncada [46] (Cuba) (Figure 1), which we studied preferably with high-resolution methodology. This selection includes the El Kef section, where the Global Boundary Stratotype Section and Point (GSSP) for the base of the Danian Stage was defined [37], and most of the designated auxiliary sections (Caravaca, Zumaia, Aïn Settara, Elles, and Bochil) [30]. The paleodepth of most of these reference sections are summarized in Molina et al. [30,37] and Schulte et al. [47]: the sections span paleodepths from outer sublittoral (El Kef, Aïn Settara and Elles) to bathyal (Caravaca, Zumaia, Agost, Gubbio, La Lajilla, Bochil, and Moncada). The biostratigraphic ranges of the planktic foraminiferal species shown in Figure 2, and the changes in relative abundance of planktic foraminiferal groups across the lower Danian shown in Figure 3, are the result of a review, compilation, and correlation of stratigraphic studies carried out by us at all these localities. The state of preservation of the planktic foraminifera in the reference ranges of the planktic foraminiferal species shown in Figure 2, and the changes in relative abundance of planktic foraminiferal groups across the lower Danian shown in Figure 3, are the result of a review, compilation, and correlation of stratigraphic studies carried out by us at all these localities. The state of preservation of the planktic foraminifera in the reference sections varies from very good (as in the El Kef section) to poor (as in the Moncada section), but in all cases are well enough preserved to permit rigorous taxonomic identification and consistent biostratigraphic and quantitative studies. The planktic foraminiferal key-biohorizons of the qualitative biozonation proposed here are the same as those used by Arenillas et al. in a previous biozonation [24]: i.e., the KPB mass extinction horizon (or the HOD of Abathomphalus mayaroensis), and the LODs of Parvularugoglobigerina longiapertura, Parvularugoglobigerina eugubina, Eoglobigerina simplicissima, Parasubbotina pseudobulloides, Subbotina triloculinoides, and Globanomalina compressa.
The key biohorizons of the proposed acme-zonation are the same as those used by Arenillas et al. [45] in defining the Planktic Foraminiferal Acme Stages (PFAS): i.e., the LODs of the Guembelitria acme (PFAS-1), the Parvularugoglobigerina-Palaeoglobigerina acme (PFAS-2), and the Woodringina-Chiloguembelina acme (PFAS-3). Here, we provide updated magnetochronological and astronomical calibrations (in kyr after the KPB) of all these keybiohorizons and an average estimate of the ages (in Ma) of the base and top of the proposed biozones.
For the magnetochronological calibrations, we rely on a recent bio-magnetostratigraphical correlation from the Caravaca section performed by Gilabert et al. [8] (see details in Supplementary Table S1). The lowermost Danian at Caravaca is characterized by a ~6 cm thick dark clay bed, with a ~2 cm thick darker clay in its basal part [4,30,48]. The base of this dark clay marks the KPB, and consists of a 1 to 2 mm thick red airfall layer containing high concentrations of iridium and impact ejecta, such as altered glass spherules, Nispinels, shocked quartz, etc. [4,[49][50][51][52]. This level coincides with the planktic foraminiferal mass extinction horizon [4,31,34,53].
To establish the age model at Caravaca, we linearly interpolate between the KPB, the top of the KPB dark clay bed, and the C29r/C29n, C29n/C28r, C28r/C28n, and C28n/C27r magnetic reversals. Following the Geological Time Scale 2020 [54] for the early Danian, which is based on astronomical calibrations [55], we assign an age of 66  The planktic foraminiferal key-biohorizons of the qualitative biozonation proposed here are the same as those used by Arenillas et al. in a previous biozonation [24]: i.e., the KPB mass extinction horizon (or the HOD of Abathomphalus mayaroensis), and the LODs of Parvularugoglobigerina longiapertura, Parvularugoglobigerina eugubina, Eoglobigerina simplicissima, Parasubbotina pseudobulloides, Subbotina triloculinoides, and Globanomalina compressa. The key biohorizons of the proposed acme-zonation are the same as those used by Arenillas et al. [45] in defining the Planktic Foraminiferal Acme Stages (PFAS): i.e., the LODs of the Guembelitria acme (PFAS-1), the Parvularugoglobigerina-Palaeoglobigerina acme (PFAS-2), and the Woodringina-Chiloguembelina acme (PFAS-3). Here, we provide updated magnetochronological and astronomical calibrations (in kyr after the KPB) of all these key-biohorizons and an average estimate of the ages (in Ma) of the base and top of the proposed biozones.
For the magnetochronological calibrations, we rely on a recent bio-magnetostratigraphical correlation from the Caravaca section performed by Gilabert et al. [8] (see details in Supplementary Table S1). The lowermost Danian at Caravaca is characterized by a~6 cm thick dark clay bed, with a~2 cm thick darker clay in its basal part [4,30,48]. The base of this dark clay marks the KPB, and consists of a 1 to 2 mm thick red airfall layer containing high concentrations of iridium and impact ejecta, such as altered glass spherules, Ni-spinels, shocked quartz, etc. [4,[49][50][51][52]. This level coincides with the planktic foraminiferal mass extinction horizon [4,31,34,53].

Taxonomic Notes
The planktic foraminiferal taxonomy used here (Figures A1-A5; Supplementary Text S1) as a basis for determining the biostratigraphic ranges of the species in Figure 2 is based on detailed morphological-ontogenetic analysis and high-resolution biostratigraphic studies mainly performed in Tunisian sections, such as El Kef [24,[37][38][39][40]. This taxonomy differs partially from the one most used by lower Danian biostratigraphers [29], both in the number of species distinguished and in the diagnostic criteria for differentiating some of the species: for example, the most widespread taxonomy [29] assigns the morphological characters of all species of Guembelitria and Chiloguembelitria ( Figure A1) to only one: Guembelitria cretacea.
Similar differences occur when distinguishing species in Parvularugoglobigerina (Figures A2 and A3; Supplementary Text S1), among which Olsson et al. [29] only consider three species (Pv. eugubina, Pv. alabamensis, and Pv. extensa); in Eoglobigerina ( Figure A4), among which they only consider two species (E. eobulloides and E. edita); and in Globanomalina ( Figure A5), among which they only consider three species (Gl. archeocompressa, Gl. planocompressa, and Gl. compressa). The different taxonomic criteria used among early Danian planktic foraminiferal specialists cause apparent differences in the stratigraphical distribution of some index species (see discussion below), and effectively compromise the rigor of the biochronological scales.
In order to provide a clearer exposition of the taxonomy followed here, we summarize in Figures A1-A5 and Supplementary Text S1 the morphological and textural criteria used to distinguish the early Danian planktic foraminiferal species and genera, and illustrate specimens of each of them, which were photographed under a Zeiss MERLIN FE-SEM at the Universidad de Zaragoza (Spain). The planktic foraminiferal specimens are from El Kef, Aïn Settara, and Sidi Nasseur (Tunisia), Gebel Aweina (Egypt), Ben Gurion (Israel), Bajada del Jagüel (Argentina), and DSDP Site 305 (Shatsky Rise, northwestern Pacific). In Figure A1, note that Chiloguembelitria exhibits a microperforate rugose and/or pustulate rugose wall texture, unlike Guembelitria, which displays a typical pore-mounded wall texture. In Figure A3, note that specimens of Parvularugoglobigerina and other parvularugoglobigerinids (Pseudocaucasina and Palaeoglobigerina) exhibit a smooth wall texture when well preserved, but a rough or microgranular wall surface when their surface is recrystallized. This contrasts with the wall texture of Trochoguembelitria ( Figure A2), which is similar to that of Chiloguembelitria. Note that the diagnostic characters of the genus Trochoguembelitria were assigned by Olsson et al. [29] to Parvularugoglobigerina. In Figure A5; note also that specimens of Acarinina trinidadensis and Acarinina uncinata exhibit a muricate wall texture when well preserved, unlike the Praemurica species.

Parallel Nomenclature in Qualitative Biozonations Based on Ranges of Index Species
The first planktic foraminiferal zonations in the 1950s and 1960s for the lower Danian used the conventional system of binomial nomenclature in naming biozones: i.e., using the name/s of index species employed to define them [1,2,10,11]. The use of a system of alphanumeric nomenclature, i.e., sequentially numbering biozones, became widespread in the 1970s [13][14][15]. Biozones with an alphanumeric nomenclature feature the disadvantage of being very inflexible and, once published, they do not lend themselves easily to the insertion of new biozones or the elimination of old ones, as these alter the order and create confusion. The same alphanumeric designation can be used in a different sense by biostratigraphers, creating added confusion. The fundamental reason is that these biozone names lack intrinsic meaning and provide very little information on the micropaleontological content.
This problem is especially relevant for the lower Danian, as at least four different alphanumeric biozonations using the "P" notation have been proposed [2,15,18,23], which is very confusing for non-specialists. However, this system provides biostratigraphers with a useful and easy mnemonic means of communication, because it automatically indicates the order and relative position of the biozones and is advantageous in both written and verbal presentation. For this reason, a combined binomial and alphanumeric nomenclature system is sufficient to resolve doubts about the alphanumeric designations.
To avoid confusion with the P-notation of previous biozonations [2,15,18,23], we here propose a new alphanumeric notation ("Dan") for the Danian biozones. Figure 2 includes a comparison of this new qualitative Dan-biozonation with the most standardized P-biozonation [2] as well as with others that have also been frequently used [18,23,24].

Inconsistencies in Qualitative Biozonations Due to Taxonomic Discrepancies
One of the main taxonomic problems in biostratigraphy is the discord between splitter and lumper taxonomists. Discrepancies over recognizing few or many species are caused by the difficulty of distinguishing "real" biological species in the micropaleontological record in the face of traditional morphological analyses (biospecies vs. morphospecies concepts). There are arguments for and against both positions. However, in practice, the species ranges proposed by two biostratigraphers cannot be accurately compared or correlated if there are taxonomic discrepancies between them, creating confusion among non-specialists.
In the lowermost Danian, the most obvious case of the divergence between splitter and lumper taxonomies is exemplified by the genus Parvularugoglobigerina [5,29,38]. According to the planktic foraminiferal taxonomy most frequently used by lower Danian biostratigraphers [29], Parvularugoglobigerina comprises only three species (Pv. eugubina, Pv. alabamensis, and Pv. extensa), with both pore-mounded and smooth wall textures. The more splitter-oriented taxonomy used here proposes up to four different genera for these morphologies: Parvularugoglobigerina, Palaeoglobigerina, Pseudocaucasina, and Trochoguembelitria, and a total of fourteen species (Figures A2 and A3). Failure to recognize the latter three genera and their species may have led some biostratigraphers to claim that the LODs of species of Globanomalina, Eoglobigerina, Praemurica, and Globoconusa are in Biozone P0 or close to the P0/Pα Biozone boundary [5,23,25,29,72].

Acme-Zonation (Quantitative Biozonation) as an Alternative
Since taxonomic discrepancies and subjectivity can greatly influence biostratigraphic studies, it is necessary to find an alternative that offers more objective criteria, even if this causes a loss of biostratigraphic resolution. For the lower Danian, this alternative is provided by quantitative data and the establishment of an acme-zonation ( Figure 3). An abundance zone or acme-zone represents the interval of maximum apogee (acme), generally the maximum relative abundance, of a particular taxon or taxon set. The boundaries of an abundance zone are defined by the biohorizons at which there is a notable change in the abundance of the index taxon or taxa (key-acme-horizons): a rapid increase for the base and a rapid decrease for the top. The acme zone takes its name from the most significant or abundant index taxon or taxa. For the definition of acme zones, it is convenient to analyze the samples quantitatively to locate their base and top with greater precision.
The use of quantitative data, especially when referring to the relative abundances of a species set (e.g., a genus or a genus set), minimizes the subjectivity and confusion present in the taxonomic determination of a particular species. Two biostratigraphers or two taxonomists may disagree when identifying a species due to their divergent preferences for splitter or lumper taxonomies or simply due to their assignation of different names. However, they are more likely to agree when identifying the acme of a species set characterized by easy-to-distinguish morphologies. For example, in the lower Danian, it is easier to agree over the recognition of an acme of parvularugoglobigerinids than in the taxonomic identification of Pv. eugubina.
A previous step in defining abundance zones is the identification of quantitative intervals or acme stages, delimited by two key acme horizons (base and top). An acme stage is an informal biostratigraphic unit that refers to each of the quantitative intervals or episodes recognizable in the stratigraphic record. To define an abundance zone, it is necessary first to trace it laterally, i.e., to check that the acme stage is useful for biochronostratigraphic correlation. An unusual relative abundance of a particular taxon or taxa in the stratigraphic record may result from a number of processes that can be local, diachronic in different localities, or repeated in different times. Acme stages with these characteristics can be useful to define ecozones, which are the minimum ecostratigraphic units characterized by shifts in type assemblages linked to paleoenvironmental changes. Their utility for correlation is normally local or regional, although it can also be global if paleoenvironmental changes are triggered by eustatic and climatic cycles (Milankovitch cycles). However, there are some acme stages linked to evolutionary processes, and these consequently do not repeat in time. If its lateral traceability or biochronostratigraphic utility is demonstrated, the identification of an acme stage may allow an abundance zone to be formally defined. Like any other biostratigraphic unit, the base of an acme zone must be based on the stratigraphic record of a bioevent that does not repeat in time.
In the lower Danian, several distinctive acme stages have been recognized [33,40,41], allowing the so-called PFASs to be identified [45]: PFAS-1 (dominance of triserial species of the genus Guembelitria), PFAS-2 (dominance of tiny trochospiral species of the genera Parvularugoglobigerina and Palaeoglobigerina), and PFAS-3 (dominance or high abundance of biserial species of the genera Woodringina and Chiloguembelina). The PFASs, which are the basis for the acme-zonation proposed here, were established after quantitative studies carried out on the >63 µm size-fraction [33,40,41,45]. This acme zonation also appears to be valid for studies carried out on <63 µm size-fractions [73,74], but the synchronicity of acme horizons recognized in these size fractions should be better contrasted. These acmestages have been identified worldwide (Figure 3), mainly in the Tethys, North Atlantic, Gulf of Mexico, and Caribbean [6,8,9,42,45,46,72,74,75] but also in the Central Pacific and South Atlantic [9,18,76], suggesting that they are useful for global stratigraphic correlation. The succession of acme stages and their synchronicity seems to be independent of the heterogeneous conditions of ocean productivity after the KPB extinction event [6,8,9,74], at least in the localities and environments where these acme stages have been recognized and analyzed by us. Figure 4 and Tables 1 and 2 summarize the age model and magnetochronological calibrations of key biohorizons in the Caravaca reference section, which are based on GTS 2020 [54] and a bio-magnetostratigraphic correlation [8,35]. Details of the magnetochronological calibration of the bases of the biozones and acme zones are shown in Supplementary  Table S1. At Caravaca, the LODs of Pv. longiapertura, Pv. eugubina, E. simplicissima, P. pseudobulloides, S. triloculinoides, and Gl. compressa are respectively placed 3, 22, 42, 107, 332, and 655 cm above the KPB (Table 2). According to the age model at Caravaca, these key-biohorizons are bio-magnetochronologically calibrated 5,19,31,68,198, and 560 kyr after the KPB, respectively (Figure 4). In addition, the LODs of the Parvularugoglobigerina-Palaeoglobigerina and Woodringina-Chiloguembelina acmes are respectively placed 5 and 55 cm above the KPB and are calibrated 8 and 38 kyr after the KPB, respectively (Figure 4).  [54] and the bio-magnetostratigraphic correlation in the Caravaca reference section [8,35]; (A)-bio-magnetochronological correlation and magnetochronologically calibrated ages of key biohorizons; (B,C)-Graphic correlations to establish the age model at Caravaca, using as tie-points (key-horizons) the KPB, the top of the KPB dark clay bed, and the magnetic reversals. Figure 5 and Tables 1-2 summarize the age model and astronomical calibrations of key biohorizons in the Zumaia reference section, which are based on the La2011 astronomical solution [69] and a bio-cyclostratigraphic correlation [9]. Details of the astronomical calibration of the bases of the biozones and acme zones are shown in Supplementary   Figure 4. Magnetochronological calibration of the bases of the biozones and acme zones proposed in this report, based on GTS 2020 [54] and the bio-magnetostratigraphic correlation in the Caravaca reference section [8,35]; (A)-biomagnetochronological correlation and magnetochronologically calibrated ages of key biohorizons; (B,C)-Graphic correlations to establish the age model at Caravaca, using as tie-points (key-horizons) the KPB, the top of the KPB dark clay bed, and the magnetic reversals. Figure 5 and Tables 1 and 2 summarize the age model and astronomical calibrations of key biohorizons in the Zumaia reference section, which are based on the La2011 astronomical solution [69] and a bio-cyclostratigraphic correlation [9]. Details of the astronomical calibration of the bases of the biozones and acme zones are shown in Supplementary Table S2. At Zumaia, the LODs of Pv. longiapertura, Pv. eugubina, E. simplicissima, P. pseudobulloides, S. triloculinoides, and Gl. compressa are placed 6, 23, 37, 100, 330, and 655 cm above the KPB (Table 2). According to the orbital tuning at Zumaia, these key-biohorizons are astronomically calibrated 7, 18, 26, 68, 210, and 473 kyr after the KPB, respectively ( Figure 5). In addition, the LODs of the Parvularugoglobigerina-Palaeoglobigerina and Woodringina-Chiloguembelina acmes are placed 6 and 55 cm above the KPB and are calibrated 7 and 42 kyr after the KPB, respectively ( Figure 5).   (Table 2). According to the orbital tuning at Zumaia, these key-biohorizons are astronomically calibrated 7, 18, 26, 68, 210, and 473 kyr after the KPB, respectively ( Figure 5). In addition, the LODs of the Parvularugoglobigerina-Palaeoglobigerina and Woodringina-Chiloguembelina acmes are placed 6 and 55 cm above the KPB and are calibrated 7 and 42 kyr after the KPB, respectively ( Figure 5).    We do not currently possess precise bio-magnetochronological and astronomical calibrations for the LOD of A. trinidadensis. At Caravaca, this key biohorizon is placed~1700 cm above the KPB [32]. According to the age model of Caravaca, the LOD of A. trinidadensis is calibrated~2113 kyr after the KPB (Figure 4; Table 2). We also do not possess detailed quantitative studies in the >63 µm size-fraction to accurately calibrate the HOD of the Woodringina-Chiloguembelina acme. Preliminary quantitative studies at some localities, such as Sidi Ziane (Algeria) and IODP Site M0077 (Chicxulub impact structure), suggest that the abundance of Woodringina and Chiloguembelina markedly decreases between the LODs of Gl. compressa and A. trinidadensis. Above this biohorizon, other genera, such as Eoglobigerina, Subbotina, Parasubbotina, Globanomalina, and Praemurica, clearly become dominant. However, this biohorizon remains vague, at least given our current state of knowledge. The abundance of these other genera increases between the LODs of S. triloculinoides and Gl. compressa, becoming codominant, or even locally or occasionally more abundant than Woodringina-Chiloguembelina (Figure 3). In the Caravaca reference section, this quantitative change, called the "lower/upper W-Ch acme" in Table 2, occurs 430 cm above the KPB, i.e., 255 kyr after the KPB according to our bio-magnetochronological calibrations (Figure 4). In the Zumaia reference-section, it occurs 365 cm above the KPB, i.e., 241 kyr after the KPB according to our astronomical calibrations ( Figure 5). Remarks: Biozone Dan1 (or the Guembelitria cretacea Zone) represents the same biostratigraphic interval as the Mh. holmdelensis Subzone of Arenillas et al. [24]. The latter used Mh. holmdelensis to name this sub-biozone, since there seemed to be strong evidence that it was a survivor of the KPB mass extinction event [6,72,77,78] and the ancestor of Danian taxa, such as Globanomalina, Eoglobigerina, and/or Praemurica [5,29,77,79] or Parvularugoglobigerina [36]. However, both phylogenetic hypotheses, and even the survival of Muricohedbergella from the KPB extinction event, have subsequently been questioned [38,39,45]. Biozone Dan1 is also equivalent to Biozone P0 of Smit et al. [4,18], but they differ in their upper boundary: the LOD of Pv. longiapertura for Dan1 and the LOD of Globigerina? minutula (probably Ps. antecessor in this paper) for P0. It could be equivalent to Biozone M18 (the Rugoglobigerina hexacamerata Zone) of Blow [15], which was defined as the interval between the HOD of Abathomphalus mayaroensis (=KPB mass extinction horizon) and the LOD of Pv. longiapertura. It is not strictly equivalent to Biozone P0 discussed by Wade et al. [2] and Keller et al. [23], because the former considered Pv. longiapertura to be a junior synonym of Pv. eugubina, and the latter regarded their LODs as synchronous. However, both biozones represent the same biostratigraphic interval as Dan1.