Bridging the extant and fossil record of planktonic foraminifera: implications for the Globigerina lineage

We conducted a morphometric study and wall texture analysis on extant and fossil specimens of the planktonic foraminifera Globigerina falconensis plexus. Our global data reveal morphological inconsistencies between fossil and extant populations. Our results are significant as G. falconensis is widely used in palaeoceanographic studies in conjunction with its sister taxon G. bulloides. Morphologically these two species are similar, with the main difference being the distinctive apertural lip present in G. falconensis. We selected cores covering the entire stratigraphic range of G. falconensis, from the early Miocene to current day, spanning sites from high latitudes in the North Atlantic Ocean and the southern Indian Ocean to sites in equatorial regions. The morphology found in the modern ocean is not consistent with the Miocene holotype of Globigerina falconensis Blow described from lower Miocene sediments in Venezuela. A more lobate morphology evolved in the late Miocene, thus, a new name is required for this morphotype, coexisting in the modern oceans with G. falconensis s.s. We thus describe the new morphospecies, G. neofalconensis for the more lobate forms which evolved in the late Miocene and inhabit the modern oceans. Additionally, we report a pseudocancellate wall texture present in the G. falconensis plexus. We use the molecular sequences from the PR2 database to explore the generic attribution of the G. falconensis lineage, confirming its close relationship with G. bulloides and its retention in the genus Globigerina.

Blow described from lower Miocene sediments in Venezuela.A more lobate morphology evolved in the late Miocene, thus, a new name is required for this morphotype, coexisting in the modern oceans with G. falconensis s.s.We thus describe the new morphospecies, G. neofalconensis for the more lobate forms which evolved in the late Miocene and inhabit the modern oceans.Additionally, we report a pseudocancellate wall texture present in the G. falconensis plexus.We use the molecular sequences from the PR 2 database to explore the generic attribution of the G. falconensis lineage, confirming its close relationship with G. bulloides and its retention in the genus Globigerina.P L A N K T O N I C foraminifera are widely used in biostratigraphy, palaeoceanography and evolutionary studies as indicators of time and past ocean conditions and chemistry.These applications require a robust identification of species, since without clear taxonomic concepts, information on diversity and palaeoecology will be misinterpreted.Taxonomy is based on types, which capture the essence of a species, but not the range of its variability (Scott 2011) and especially not through time.This often leads to conflicting classifications and in some cases, the established concept of a morphospecies may even become disconnected from the type specimen (Fabbrini et al. 2021).For example, this could occur when a species is associated with a type that has been considered in isolation of the variability in the population from which it is derived (Scott 2011), as may happen when scanning electron micrograph (SEM) images of the type specimen are unavailable.Every situation where the taxon concept is not consistent with the type causes taxonomic instability (Wade et al. 2018;Fabbrini et al. 2021).In organisms that have a rich fossil record, this problem is epitomized when names of living taxa are based on fossil types.It is extremely important to understand the living plankton and to put the modern taxa in the context of their taxonomic history and fossil record (Morard et al. 2022).A taxonomic review of the living planktonic foraminifera was undertaken by Brummer & Ku cera (2022), in which they recognized 50 extant species living in the modern oceans.Planktonic foraminifera are an example of a group in which the palaeontological and biological classifications are interwoven, and two extant species have fossil types: Globoquadrina conglomerata and Globigerina falconensis.
Brummer & Ku cera (2022) highlighted the issues that arise when extant species are described from sediments rather than from the plankton, and the need for a thorough assessment to demonstrate the equivalency of fossil types and their living species.This is the case for the living taxon G. falconensis (Fig. 1), a long ranging species described by Blow (1959) from the Burdigalian (lower Miocene) sediments of Falc on (Venezuela).
The genus Globigerina is typified by trochospiral coiling, globular chambers, a single umbilical aperture and spinose wall consisting of small (<0.9 lm in diameter) pores and spine collars, termed bulloides-type by Hemleben & Olsson (2006).It ranges from the middle Eocene (Zone E9) to today.There are several fossil genera sharing morphological similarities with Globigerina, but they all lack the characteristic bulloides-type wall; for example, Subbotina and Globoturborotalita, which have a cancellate wall texture (Hemleben & Olsson 2006).The current classification of the genus Globigerina comprises two extant species, G. bulloides and G. falconensis (Fig. 1), which are common in modern assemblages from many marine environments.The two species are morphologically similar, both possessing four globular chambers in the final whorl, increasing slowly in size, with an umbilical aperture.The chief distinction between G. falconensis and G. bulloides is the characteristic apertural lip present in G. falconensis.The two taxa show some ecological differences, and for this reason they are commonly used in palaeoclimatic studies (Malmgren & Kennett 1978;Naidu & Malmgren 1996;Li et al. 2001;Xu et al. 2005).Globigerina bulloides is most abundant at higher latitudes and in eutrophic areas, such as upwelling localities.The extant G. falconensis has a more cosmopolitan range (Siccha & Kucera 2017) and is more abundant in mid-latitudes, and in monsoonal and upwelling conditions (Malmgren & Kennett 1978;Naidu & Malmgren 1996;Li et al. 2001;Xu et al. 2005).Both G. falconensis and G. bulloides inhabit the mixed-layer (Sousa et al. 2014; among others).The correct identification of G. falconensis is then crucial to understand the extent and origin of biodiversity in living planktonic foraminifera, to interpret the evolution of the group, and to constrain its palaeoecological applications.Globigerina falconensis has been reported frequently in the fossil record, and its lowest occurrence has been used to calibrate molecular clocks in genetic studies (Andr e et al. 2014;among others).
Despite clear evidence for morphological distinction, G. falconensis has been misidentified or grouped with its sister taxon G. bulloides (Al-Sabouni et al. 2018;Fenton et al. 2018;Hsiang et al. 2019), and some authors have even questioned the validity of G. falconensis (B e 1968;Kennett 1969).To solve this issue, morphological analyses were conducted in order to define the key features distinguishing G. falconensis from G. bulloides (Malmgren & Kennett 1977), but that study neglected the extensive analysis of fossil specimens.Therefore, we revisited the taxonomic concept of G. falconensis by sampling and analysing large populations of G. falconensis from multiple locations and throughout its stratigraphic record, from the early Miocene to the current day.Here we sought to establish whether there is consistency between fossil type and living forms of G. falconensis through the study of both fossil and extant populations, including the reexamination of the type material deposited at the Natural History Museum (London).the fossil range of G. falconensis globally and with good fossil preservation, spanning from two sites at high latitudes (ODP Site 982 in the North Atlantic Ocean and ODP Site 747 in the southern Indian Ocean) to two sites in equatorial regions (IODP Site U1489 and ODP Site 871), and the Meteor M32 in the subtropical Arabian Sea.ODP Site 925 is located in the western equatorial Atlantic, and constitutes a particularly relevant study area, since it is not far from the type locality of G. falconensis Blow, 1959, in Venezuela.

Micropalaeontological analysis
Sample preparation.All samples were prepared following the standard washing, drying and sieving procedures.First, the samples were pre-soaked in c. 150 mL of distillate water to disaggregate the sediments, and then washed with tap and distilled water, sieving the sediment through a 63 lm mesh.The residues were oven dried at 40°C.Different size fractions were obtained through 250 and 125 lm dry sieves for each sample.The samples from Meteor M32/2 (Recent from the Arabian Sea) were sieved in different size fractions: above 400, 400-315, 315-250, 250-200, 200-150, below 150 lm.Total population picking was conducted in order to characterize the assemblages.Microfossil specimens were examined under a continuous-zoom stereo microscope.Well preserved specimens and entire populations available of the G. falconensis plexus were picked and stored on microslides for further imaging and analysis.Species identification was based on the literature (Bolli 1957;Blow 1959Blow , 1969;;Kennett & Srinivasan 1983;Bolli et al. 1985;Spezzaferri 1994;Aze et al. 2011;Fox & Wade 2013;Wade et al. 2018) and the online archive Mikrotax (Huber et al. 2016).
Imaging.Unbroken specimens were selected for SEM imaging and further scanning.The selected specimens were stuck on metal stubs using double-sided sticky tape.The stubs were coated with gold and inspected using a Jeol JSM-6480LV high-performance, variable pressure analytical scanning electron microscope at the Department of Earth Sciences, University College London (UCL).Optical photographs were taken using an OLYM-PUS DP73 multifocal camera mounted on an OLYMPUS SZX16 stereo microscope at the at the Department of Earth Sciences, UCL.The images were taken using a step of 15 lm and an average imaging time of 60 s per specimen.The images were then postprocessed using the software Stream Motion (Olympus).
Micro computed tomography.One specimen was selected for a high-resolution imaging using micro computed tomography (micro-CT).The three-dimensional microstructure of the selected specimen was studied using a ZEISS Xradia 620 Versa x-ray computed tomography microscope, located at the Electrochemical Innovation Lab (EIL), part of the Department of Chemical Engineering, UCL.The specimen was mounted as single specimen set up following a protocol modified after Coletti et al. (2018).The sample was scanned for a total of 60 min with a maximum resolution of 500 nm.In total, 1601 radiographs were acquired over a 360°sample rotation range with an exposure time of 25 s per radiograph.The sample was placed between the x-ray source and a 2 k 9 2 k detector with a source-to-detector distance of 39.9 mm providing a voxel resolution of c. 500 nm using the 209 objective magnification in binning 1 mode.The instrument was operated at 80 kV and 7 W, employing a low energy filter to optimize transmission and contrast to noise ratio.The raw transmission images from micro-CT imaging experiment were reconstructed using a commercial image reconstruction software package (Zeiss XMReconstructor, Carl Zeiss X-ray Microscopy Inc., Pleasanton CA), which employs a filtered back-projection algorithm.The reconstructed greyscale 3D image volumes were subsequently segmented using the Avizo 3D 2022.1 software package (ThermoFisher Scientific, Waltham MA).The surface was then generated and saved as an STL file.
Morphometry.All measurements for the morphometric study were conducted on the entire picked populations of G. falconensis plexus, using the software Image Pro, and ana QImaging RETIGA-2000R camera mounted on a light microscope at the Department of Earth Sciences, UCL.The following three parameters were selected and measured: area (A) (lm 2 ), diameter (D) (lm), roundness (R).This last parameter (roundness) was obtained from A and D, using Image Pro.All data were plotted using the statistical software PAST (Hammer et al. 2001).
Molecular analysis.To investigate the genetic evolution of the G. falconensis plexus, we accessed curated SSU rDNA sequences of the species G. falconensis, Globoturborotalita rubescens, Globigerinoides tenellus, Globigerinoides elongatus, Globigerinoides conglobatus and Globigerinoides ruber albus from the PR 2 database (https://pr2-database.org;Guillou et al. 2013) (Table 1) using the R package PR 2 database in R v4.1.1 (R Core Team 2013; Vaulot 2022).To ensure that the resulting topology is not affected by long-branch attraction towards G. bulloides and the phylogenetic placement of G. falconensis emerges independently, we excluded on purpose G. bulloides.We set up the molecular clock analyses using the divergence between Globoturborotalita rubescens (as representative of the entire genus Globoturborotalita) and Globigerinoides at We used a 'relaxed' clock model as implemented in BEAST v1.8.4 (Suchard et al. 2018).The model parameters were set using BEAUti v1.8.4.The distribution of the fixed node age prior was considered normal and the speciation rate was assumed constant under the Yule-Process.The generalized time reversible (GTR) model was selected as a substitution model.Markov-Chain Monte Carlo (MCMC) analyses were conducted for 10 000 000 generations, with a burn-in of 1000 generations and saving each 1000th generation.The maximum clade credibility tree with median node heights was calculated in TREEAnnotator v1.8.4 from the BEAST package, with a burn-in of 100 trees and a posterior probability limit of 0. The resulting tree was then imported into the R environment using the read.beastfunction from the treeio package (Wang et al. 2020).The tree was successively visualized using the R package ggtree (Yu et al. 2017) and a geological time scale was added using the R package deeptime (Gaerty 2022).

RESULTS
We succeeded in extracting 260 specimens belonging to the G. falconensis plexus from samples ranging from Zone M5 (sensu Wade et al. 2011) to Recent from multiple global locations (Fig. 2).The specimens of the G. falconensis plexus could always be distinguished from G. bulloides by the distinctive apertural lip.During our examination we noted a difference in the overall shape of the test of the recent specimens.While Miocene populations possess a morphology consistent with the holotype designated by Blow (1959), the Pliocene and Quaternary populations, in contrast, exhibit a more lobate outline in umbilical view, with a tendency for chamber elongation and a flatter spiral side.These characters are not present in the Miocene specimens, deviating from the holotype of G. falconensis.Globigerina falconensis s.s is still present in modern assemblages, but the overall number of such individuals is lower than that of lobulate individuals.
Supporting these preliminary observations, the key parameters we measured have different values (e.g.roundness, or the lobateness of the outline) through the fossil record.In Figure 3, the distribution of two parameters (roundness and maximum diameter) shows a marked shift in the occupancy of the morphospace through time, with the Recent populations (average roundness = 1.21) being significantly different from the Miocene populations (average roundness = 1.12).This difference is not related to the size of the specimens Accession number of the SSU rDNA sequences extracted from the PR 2 database (Guillou et al. 2013) used for the molecular clock analysis (Fig. 6).

Accession no.
Genus Species KM194166 Globigerinoides conglobatus analysed (Fig. 3).The change through time is such that the Recent specimens plot outside of the range of variability recorded in the Miocene populations.Pliocene and Pleistocene populations still retain individuals comparable to G. falconensis s.s.but the frequency of the lobulate types is consistently higher (Fig. 3).

Wall texture
While investigating the lineage morphology, we noted unexpected differences and variability in the wall texture of the specimens.Comparison of the different Recent G. falconensis populations investigated globally and G. bulloides specimens are illustrated Figure 4. Normally the wall texture of G. falconensis is reported in the literature as bulloides-type (Fig. 4), as in its ancestor species G. bulloides.We measured the pore number and mean diameter in a standard area of 50 9 50 lm.Our data indicate a difference between the G. falconensis plexus and the other globigerinids (Fig. 5).The average measured pore diameter is 1.7 lm, but the size interval spans from 3.0 to 1.0 lm.The total number of pores in the standard unit areas ranges from 20 to 55, with 32 as the mean value.The range of the G. falconensis plexus overlaps the 'morphospace' typical of Globoturborotalita, with slightly larger pores than G. bulloides (Fig. 5).No substantial differences emerged between modern and fossil specimens of the G. falconensis plexus (Fig. 5), which showed consistent values through space and time.Modern specimens from the Atlantic Ocean at different latitudes have been imaged for comparison with the Arabian Sea populations (Meteor M32/2).These specimens are shown in Figure 4 together with G. bulloides from the same assemblages, in order to document the difference in wall texture between the two taxa.The main discriminant between the two morphospecies is the development or partial development of inter pore ridges in the G. falconensis plexus.We refer to this wall texture as pseudocancellate.The development of interpore ridges is absent in G. bulloides, and it is indeed considered a key feature of other genera, such as Globoturborotalita, Globigerinoides and Trilobatus.Therefore, we felt compelled to re-examine the phylogenetic position of the G. falconensis lineage using the modern molecular genetic sequences available as an independent source of information.

Genetic results
We selected SSU rDNA sequence extracted from single cells and attributed to G. falconensis collected in the Great Barrier Reefs, and available in the PR 2 database (Guillou et al. 2013).The topology obtained shows the phyletic relation between G. falconensis and the other genera of Globigerinidae, such as Globoturborotalita, and Globigerinoides (Fig. 6).
The resulting tree places G. falconensis as a distinct clade to Globigerinoides and Globoturborotalita rubescens.The molecular clock calibrated on divergences within Globigerinoides and Globoturborotalita rubescens indicates an ancient split between the G. falconensis plexus and Globoturborotalita estimated at around 29 Ma, in the early Oligocene.This split is then much older than the complete fossil record of G. falconensis, which appeared around 17 Ma (early Miocene).This datum demonstrates that G. falconensis and Globoturborotalita have a significant degree of genetic divergence, indicating that they are not closely related.Additionally, this suggests that G. falconensis belongs to the genus Globigerina with its sister taxon and ancestor G. bulloides.

Globigerina falconensis plexus
We aimed to link the extant and fossil record of the G. falconensis plexus and use morphometric data to determine the variability of the plexus and link the holotype of G. falconensis, (early Miocene) with the modern populations.Both visually and quantified by our morphometric analysis, there is a difference in the shape of the test in the Recent specimens, which exhibit a more lobate outline, with a tendency for chamber elongation and a lower trochospire.Our morphometric and genetic data thus present a conundrum.There are two morphotypes in the modern ocean, both possessing four chambers in the final whorl, and a lip, but they are distinguished by the periphery.The lobulate morphospecies is inconsistent with the holotype of G. falconensis, which is much more compact (Fig. 7).The more lobate morphology is not found in the fossil record in sediments before the upper Miocene, while the compact G. falconensis s.s.persists up to the modern.There are specimens that can be attributed to G. falconensis s.s. in Recent sediments, but these are smaller in size and rare.Our morphometric and imaging study suggests that there  are two types of G. falconensis extant in the modern ocean, one with a lobate periphery which evolved in the late Miocene, and G. falconensis s.s. which evolved in the early Miocene.The holotype of G. falconensis is representative of the variability observed in the Miocene, thus we suggest that a new species and concept is needed for the late Miocene to Recent representatives of the lineage.Therefore, to bridge the extant and fossil record and resolve the current inconsistency between the G. falconensis holotype and the modern morphotype, we describe the new morphospecies G. neofalconensis.Using mean values from the biometric parameters among the best-preserved modern specimens, we selected the type specimens (Figs 7, 8) for the new taxon (see Systematic Palaeontology, below).
Our observations indicate that the modern lobate and flatter morphology emerged in Zone M13 (late Miocene) (Fig. 9).The values extrapolated also allowed a global comparison, showing how in all sites, the lobate forms of the falconensis plexus were the most abundant in Recent sediments and consistently different from G. bulloides.In all the Recent sediments studied, a minority of G. falconensis specimens with a compact test and pseudocancellate wall texture were present in the smaller size fraction (<150 lm), indicating that G. falconensis s.s.remains extant but rare.Most specimens from sediments younger than Pliocene age are lobate, and consistent with G. neofalconensis.

Bringing together genetic and imaging analyses
The conundrum concerning the lobate morphology is then solved with the identification in the fossil record of G. neofalconensis after the late Miocene.The wall texture inconsistency still requires some further discussion.Globigerina falconensis has always been considered to be related to G. bulloides (Blow 1959(Blow , 1969;;Kennett & Srinivasan 1983;Bolli et al. 1985; among others) and part of the genus Globigerina, sharing the same type of wall texture.Specimens of G. bulloides showing their characteristic wall texture of small and irregularly distributed pores, lacking prominent inter-pore ridges are illustrated in Figure 4. Our analysis and measurements on wall texture from specimens of the G. falconensis plexus ranging from the early Miocene to the Recent, indicate a high degree of variability (Fig. 4).In several instances in both fossil and modern populations, we find specimens with a pseudocancellate appearance in the earlier chambers, developing into more of a bulloidestype wall texture in the final and penultimate chambers.The SEM and micro-CT images (Figs 4, 7) reveal that this pseudocancellate appearance is not honeycomb-like as in the sacculifer-type wall texture, typified in the modern Trilobatus sacculifer and many fossil species (Hemleben & Olsson 2006), but the pseudocancellate texture results from larger pores and interconnecting ridges (Figs 4, 7).
The molecular clock calibrated age of the split between G. falconensis and Globigerinoides-Globoturborotalita  The genus Globigerina d'Orbigny, 1826, has been amended several times considering different criteria, from the morphology to wall texture (Bolli 1957;Blow 1959;Kennett & Srinivasan 1983;Bolli et al. 1985;Spezzaferri 1994;Pearson et al. 2006;Spezzaferri et al. 2018;among others).The milestone works of B e (1968) and   Fleisher (1974) changed the taxonomy of planktonic foraminifera radically, making characters of the test wall (pore size and wall features) significant for the first time.
Following the taxonomical guidelines applied in the recent literature, wall texture is the key feature used to classify planktonic foraminifera (B e 1968;Kennett & Srinivasan 1983;Bolli et al. 1985;Olsson et al. 1999;Pearson et al. 2006;Aze et al. 2011;Wade et al. 2018).The holotypes of both G. falconensis and G. neofalconensis (Fig. 7) present the same pseudocancellate wall texture not consistent with the current definition of the genus Globigerina (B e 1968;Kennett & Srinivasan 1983;Bolli et al. 1985;Olsson et al. 1999;Pearson et al. 2006;Wade et al. 2018).Either the pseudocancellate wall texture evolved independently twice in the lineage, complicating the definition of Globigerina or this character remerged from the genetic past of the genus.In this context, the apparently pseudocancellate wall in G. falconensis lineage (Fig. 4) must fall within the range of variability of the Globigerina-type wall.Further studies are necessary to understand how this wall type can develop.A possible answer might arise from the genetic history of the genus Globigerina.
The earliest Globigerinidae also had a cancellate wall (Eoglobigerina in the Danian, early Paleocene).Such wall texture diversified during the radiation of the clade until the bulloides-type wall emerged in the Eocene from the cancellate genus Subbotina.Although Globigerina officinalis is the first member of its genus in the middle Eocene, the first occurrence of the true bulloides-type wall predates this species in its ancestor Subbotina crociapertura or Subbotina roesnaesensis.These taxa were retained by Olsson et al. (2006) in Subbotina despite already having the Globigerina-type wall texture.Moreover, when the genus Globigerina emerged, the wall texture was still variable as shown by individuals of G. officinalis presenting different types of wall texture on different portions of their test (Olsson et al. 2006;Spezzaferri et al. 2018).Eventually with the appearance of G. archaeobulloides and G. bulloides the wall texture became consistently bulloidestype.Since the G. falconensis lineage belongs to the same clade as G. bulloides, the pseudocancellate wall texture must have re-emerged in the Miocene.The phenotypic plasticity of planktonic foraminifera might allow similar wall textures to emerge repeatedly through time (Kendall et al. 2020).Alternatively, if the emergent character is similar to the plesiomorphic state, all descendants could retain it.Thus, all Globigerinidae might have retained the ability to build a pseudocancellate wall, suppressed in certain species like G. bulloides, and expressed in others as in the G. falconensis lineage.The specimens with a mixed wall texture could indeed demonstrate plasticity during ontogenetic development expressing traits from the genetic past of the lineage.These arguments would benefit from focused research on both modern and fossil planktonic foraminifera.
Molecular data indicate that there is only one form of G. falconensis in the modern ocean that is closely related to G. bulloides (e.g.Brummer & Ku cera 2022).Thus, the description of a new morphospecies, G. neofalconensis, aligns with the morphological but not with the genetic evidence.A similar situation is not new in planktonic foraminifera and it finds its best example in T. sacculifer plexus, which consists of four extant morphospecies T. sacculifer (Brady 1877), T. quadrilobatus (d'Orbigny 1846), T. immaturus (LeRoy 1939) and T. trilobus (Reuss 1850).Each of these morphospecies has a different biogeography and stratigraphic history (Poole 2017), but molecular genetic and culturing studies of extant specimens (e.g.Hemleben et al. 1987;Andr e et al. 2014) suggests that all four morphospecies belong to the same biological species.The variation between the morphospecies within the T. sacculifer plexus is considered to be ecophenotypically controlled (e.g.Hecht & Savin 1972;Hecht 1974;Andr e et al. 2014;Schmidt et al. 2016).Morphological studies on fossil specimens (Poole & Wade 2019) support the hypothesis that T. sacculifer plexus morphospecies are the same biological species, but highlight the necessity to retain the morphospecies in order to increase their biostratigraphical and palaeoecological value.We applied here the same principle to the G. falconensis plexus, to retain as much information as possible through its stratigraphic record and evolution.Size.Maximum diameter of the holotype: 320 lm.

SYSTEMATIC PALAEONTOLOGY
Remarks.Distinguished from G. neofalconensis by a closer umbilicus, a more compact test due to the tighter coiling, and with less incised sutures.Globigerina falconensis can be distinguished from G. bulloides by a well-developed apertural lip, narrower, lower-arched aperture, and the presence of a cancellate honeycomb wall texture.The coiling direction is reported in the literature as random (Blow 1959(Blow , 1969)), but based on our observation the two subspecies tend to have a different preferential coiling direction.Within our dataset, G. falconensis is predominantly left coiling, while G. neofalconensis has a random coiling.
Globigerina falconensis shows commonalities with some taxa belonging to the genus Globoturborotalita bearing four chambers in the final whorl and an apertural lip or rim, such as Gt.ouachitaensis, Gt. druryi, Gt. occlusa and Gt.eolabiacrassata.Globigerina falconensis can be differentiated from Gt. ouachitaensis due to its thicker and more developed apertural lip, the lower-arched aperture and the overall compact morphology and commonly larger size.Globoturborotalita druryi is characterized by a thicker apertural lip, larger size than G. falconensis, a more compact test outline and a higher trochospire visible when observed in edge view.Globigerina falconensis is distinguished from Globoturborotalita occlusa by the consistently developed apertural lip, which is sometimes absent in Gt. occlusa, and by its more compact shape, while Gt.occlusa tends to be lobulate and present more incised sutures with spherical chambers.Globoturborotalita eolabiacrassata can be separated from G. falconensis by tighter coiling, smaller size, more coarsely perforated test and a shorter-arched umbilical aperture, bordered by a thick imperforated rim.
visibly variable pore size.The penultimate chamber is commonly pseudocancellate, the third ultimate can instead present a Globigerina-type wall.Sometimes specimens with completely pseudocancellate wall occur, showing that the ontogenetic pattern is inconstant.
Test morphology: Test loosely coiled constituted by four subspherical slowly increasing chambers in the final whorl.The outline is lobate.A total of 10-12 chambers are arranged in 2-2.5 whorls, with random coiling direction.In umbilical view, the last whorl consists of subspherical chambers loosely coiled and separated by straight and incised sutures.The shape of the last chamber is quite variable, sometimes developing an elongated bulb-like shape, though kummerforms are very common.
Umbilicus deep and open, a low umbilical aperture bordered with an evident imperforate lip.The aperture can create a curved shape of the lip.In spiral view, four subspherical chambers divided by straight incised sutures, all the chambers from the previous whorls are visible in the spire.In edge view, low trochospire and flat spiral side with the last chamber sometimes slightly tilted towards the umbilicus, margin rounded.
Size.The maximum diameter of the holotype is 340 lm.
Remarks.This new species is named after its ancestor G. falconensis, to retain a well-established and widely used taxonomical name and concept.Globigerina neofalconensis can be distinguished from its ancestor G. falconensis by its more lobate profile and a more loosely coiled test, coupled with a wider umbilicus (see Table 2; Figs 7, 8).Blow (1969) reported two different types of G. falconensis showing distinct stratigraphic ranges, named G.falconensis forma typica and G. falconensis forma atypica.The new species presented here matches with the latter, in terms of the overall morphology and stratigraphic range.
Globigerina neofalconensis can possess both a pseudocancellate and bulloides-type wall texture.The variability in the wall texture in the different chambers as added is not always evident, making its recognition complex.Specimens of G. neofalconensis with a fully developed bulloides-type wall texture have been reported by various authors (Br€ onnimann & Resig 1971;Srinivasan 1975;B e et al. 1977;Malmgren & Kennett 1977;Thunnell 1979;Jenkins et al. 1986;Hemleben et al. 1989;Iaccarino & Salvatorini 1979;Dowsett & Robinson 2007;Parcerisa et al. 2008;Lam & Leckie 2020;Schiebel & Hemleben 2017).Culture studies have indicated that pore size can be plastic and influenced by temperature and metabolic rate (Burke et al. 2018) and thus some of the variability in wall texture appearance may be due to environmental influences.
Globigerina neofalconensis is distinguished from G. bulloides by its apertural lip, the low aperture and the very lobate profile.Commonly the last chamber of G. neofalconsis can be kummerform or radially elongated and bulb shaped.Both of these features are absent or rare in G. bulloides.Globigerina neofalconensis can be differentiated from G. antarctica due to its coarser and hispid wall texture, the well-developed apertural lip and in having four chambers in the last whorl, while sporadically individuals of G. antarctica may have five chambers in the final whorl, they are always characterized by a thin wall texture and a very thin lip.
Globigerina neofalconensis can be compared with various species within Globoturborotalita, such as, Gt. foliata, Gt. ouachitaensis, Gt. pseudopraebulloides and Gt.occlusa.Globigerina neofalconensis can be distinguished from Gt. foliata because of its apertural lip and wider umbilicus.Globigerina neofalconensis differs from Gt. ouachitaensis by its bigger size, more open umbilicus and interomarginal aperture bordered by a thicker lip.Globigerina neofalconensis can be also distinguished from Gt. pseudopraebulloides thanks to the apertural lip, and the lower aperture and flatter spiral side.Globigerina neofalconensis differs from Gt. occlusa in having a longer aperture bordered by a lip and having a looser coiling.
In the literature, there are several examples of attempted reclassification of G. falconensis.Br€ onnimann & Resig (1971) described different globigerinids belonging to the G. falconensis group.They reported the presence of different morphotypes under the name G. falconensis and suggested the possibility of a group of taxa that are quasi-homeomorphic to G. falconensis.Anomalies were noticed in the wall texture and used as a possible base to untangle the controversy (Br€ onnimann & Resig 1971).We consider all the specimens to be G. neofalconensis, despite the different wall texture in the individuals shown in this paper.However, we retain in G. falconensis all the specimens described by Br€ onnimann & Resig (1971) with a tighter coiling, more compact and thicker test.We also consider Globorotalia palpebra (Br€ onnimann & Resig 1971) to be taxon inquirendum, due to the lack of information, and its absence from other studies published in literature, and its overall rarity.

CONCLUSION
We have endeavoured to link the extant and fossil records of Globigerina, focusing on the species G. falconensis described from early Miocene sediments, and still alive in the oceans today.We investigated specimens of the G. falconensis plexus from the modern and fossil record with an integrated approach.Our morphometric and microscopy studies reveal that the majority of extant forms commonly found in the modern oceans are more lobate, and loosely coiled, and not consistent with the fossil holotype of G. falconensis.We therefore name a new morphospecies G. neofalconensis to resolve the conflict between the type specimen of G. falconensis and the living populations.
The new morphospecies G. neofalconensis, evolved in the late Miocene and inhabits the present oceans coexisting with rarer populations of G. falconensis s.s.Scanning electron microscopy highlighted how this entire plexus presents a peculiar wall texture, different to the typical bulloides-type, which characterizes the genus Globigerina.This apparent inconsistency of the wall texture led us to approach the lineage from a genetic perspective.Our molecular analysis indicates that G. falconensis and Globorturborotalita lineages split at an estimated 29 Ma, in the Oligocene.This age predates the existence of G. falconensis in the fossil record, and thus excludes any direct phylogenetic link between this species and the Globoturborotalita plexus, supporting the hypothesis that the two are clearly separated lineages.All these data allowed us to retain the new morphospecies G. neofalconensis within the genus Globigerina.

F
I G . 3 .On the left, histograms represent the temporal shift in the mean values of the roundness (R), highlighting the development of a progressively more lobate test in the G. falconensis plexus after the late Miocene.The mean value of R in the Miocene specimens is 1.12, and it reaches 1.20 during the Pliocene to Recent interval.Modern mean value (R = 1.20) is indicated by the dotted vertical line.Miocene specimens are represented in blue, Plio-Pleistocene ones in yellow and Recent specimens in red.The outlines of the two morphotypes are drawn at the top of each histogram illustrating the differences in roundness.The cross plot on the right shows the measured values of diameter (D) vs roundness (R).The specimens are represented by different symbols according to their stratigraphic age: Miocene with blue empty squares, Plio-Pleistocene with yellow squares and Recent with red circles.These plots show how R is the key parameter describing the morphological evolution through time of the G. falconensis plexus.

F
I G . 5 .Wall texture diagram showing the concentration of pores per 50 9 50 lm surface area (modified after B e 1968).Globigerina falconensis plexus and has similar values to G. bulloides.The same colour code from the previous figures is applied here (Miocene in blue, Plio-Pleistocene in yellow and Recent in red).

F
I G .6 .Molecular clock estimates of the diversification of the Globigerina falconensis plexus from the genera Globoturborotalita and Globigerinoides.The genera separated in the early Oligocene (28.93 Ma), before the appearance of G. falconensis in the fossil record, neglecting any direct phylogenetic link between the two lineages.The pink bars indicate the uncertainties in the dating each node; grey shading at tips represents the intraspecific variability.F I G .7 .Scanning electron and optical micrographs of holotype specimens: A-F, Globigerina neofalconensis sp.nov.NHMUK ZF 9958 (holotype), M32/2MC6, Arabian Sea, Holocene (Zone PT1): A-C, micro-CT images; D-F, optical micrographs.G-L, G. falconensis Blow, 1959, USNM MO 625697 (holotype; Smithsonian Institution), Falc on (Venezuela), early Miocene (Zone M5): G-I, SEM micrographs; J-L, optical micrographs.Scale bars represent 100 lm.

(
Fig. 6) shows the ancient separation of these groups and thus confirms the higher genetic similarity of G. falconensis with G. bulloides (Stewart et al. 2001).These observations indicate the G. falconensis lineage must be retained in the genus Globigerina despite the difference in wall texture.

F
I G . 9 .Synthesis of the morphological evolution of Globigerina falconensis plexus.Each group of specimens is correlated to its corresponding zone (Wade et al. 2011).The phylogenetic relationship between the G. falconensis group and G. bulloides is shown on the right.The holotypes of G. falconensis and G. neofalconensis sp.nov.are illustrated inside the boxes on the side as a reference.All presented specimens are among the individuals measured for the biometric study herein (Zone M6-M11 specimens from ODP Site 590; Zone M12-PT1 from IODP Site U1482; Recent specimens from M32/2MC6).All scale bars represent 200 lm.

F
A B B R I N I E T A L .: I M P L I C A T I O N S F O R T H E G L O B I G E R I N A L I N E A G E 1 1 Institutional abbreviation.NHMUK, Natural History Museum, London, UK.Remarks.The first representative of the genus is G. officinalis, emerging in middle Eocene Zone E10 (Olsson et al. 2006).The genus diversified in the Oligocene (Wade et al. 2018).Globigerina bollii Cita & Premoli Silva, fig.1a-c.non 1972 Globigerina antarctica Keany & Kennett, fig. 6. ?1969 Globigerina nilotica Viotti & Mansour, fig.1a-c.Test morphology.Wall spinose, normal perforate, mean pore size 1.7 lm, mean pore concentration 32 per unit area (2500 lm 2 ).The wall texture is variable ranging from a bulloides-type pitted surface to a pseudocancellate texture, with coalescing ridges arranged around the pores, creating a honeycomb appearance (Figs 4, 7).Test low trochospiral with ten to twelve chambers arranged in about two whorls and with four subspherical chambers in the last whorl.Random coiling, but with a tendency for predominately sinistral.Subspherical chambers, slightly embracing, especially the last chamber, increasing regularly and slowly in size as added, and separated by slightly incised straight sutures.The umbilicus is small and deep, sometimes almost closed by the strongly developed lip on the last chamber.The umbilical aperture is an elongate low arch, sometimes slit-like, almost straight, with a well-developed imperforate lip.Umbilical sutures are straight and radial, but not strongly incised.In spiral view, all chambers of the trochospire are visible, subspherical and separated by radial and straight to slightly curved sutures, weakly incised.