A replacement neotype for Globigerina crassaformis Galloway & Wissler, 1927, is proposed for one that was published but lost in curation prior to being deposited. The specimen is selected from an analysis of the axial shape of 54 specimens from the type and adjacent localities. Its cone-like attributes include progressively elevated and tapering umbilical walls, an acute, keeled periphery, and a low-relief spiral surface. Globorotalia crassacarina Scott, Bishop & Burt, 1990, and possibly Globorotalia hessi Bolli & Premoli Silva, 1973, include specimens that closely resemble the proposed neotype and may be junior synonyms.

Following advice from Prof. Ellen Thomas that the holotype of Globigerina crassaformisGalloway & Wissler, 1927, had been destroyed, Scott (2019) proposed a specimen from the type locality (Quaternary) at Lomita Quarry, Los Angeles County, California. This specimen was lost in curation and was not deposited. The aim of this study is to identify a replacement neotype from the type locality that considers the attributes of the holotype and complies with the articles of the International Code of Zoological Nomenclature (ICZN). It differs from Scott’s (2019) procedure, which attempted to preserve recent usage (e.g., Schiebel & Hemleben, 2017, fig. 2). Features of the holotype are re-assessed, and a larger sample of 54 specimens from the type locality and nearby collections from Lomita Marl is analyzed. The proposed neotype is cone-like with a low-relief spiral face and strongly elevated umbilical walls that taper towards the umbilicus. A low keel borders much of its subquadrate spiral outline.

Scott et al. (2015) illustrated 25 specimens found in washed residue provided by James Kennett from samples collected by Orville Bandy. The >149-μm fraction was searched for planktonic specimens with about four chambers in the outer whorl and a narrow, rather slit-like extra-umbilical aperture. These criteria allowed for inclusion of taxa that resemble Truncorotalia crassaformis (e.g., Globorotalia crassulaCushman & Stewart, 1930). The current study incorporates 29 specimens supplied by James Ingle from Lomita Quarry and nearby localities; how they were identified is unknown. Lomita Marl represents a sheltered shelfal or bank-top environment, whereas the species has been modelled as inhabiting the mesopelagic zone (Boscolo-Galazzo et al., 2022). Many specimens are damaged and probably re-deposited. Morphometric data acquisition and processing follows Scott et al. (2015) and Scott (2019). Appendix 1 lists the sources of specimens 1–54 used in the analyses (Fig. 1).

The outline of specimens in axial orientation was captured from a scanning electron microscope (SEM) and print imagery using tpsdig2 (https://www.sbmorphometrics.org/) and processed with the Procrustes package shapes (https://cran.r-project.org/web/packages/) to align specimens on their centroids after removing effects of size and position (Appendices 1–3). Deviations in axial shape are presented on the largest axes of variation resolved by principal component analysis of the Procrustes residuals matrix. Grice & Assad (2009) provide a detailed example of the algorithms. Unsupervised clustering (Ferraro et al., 2019) is used to assess the presence of shape groups within these data.

Scott et al. (2015) found that disparity in axial shape was high, extending from cone-like (ventroconical) to biconvex forms. Expectedly, this finding is confirmed with the extended dataset, which includes Pliocene material from the Repetto Formation (Appendix 1). Maximum disparity in shape is represented by specimens that map onto the convex hull of the dataset (Fig. 1A). Many are ventroconical and are furthest from the biconvex specimens which, except for specimen 49 (Fig. 1A), are kummerforms (last-formed chamber is smaller than its predecessor). Machine learning via fuzzy clustering identified five groups; ventroconical specimens (Fig. 1B, orange color) are identified with highest confidence. Many specimens in the remaining groups, which are closely associated in shape space, are identified with very low confidence (Fig. 1C).

Order Foraminiferida Eichwald, 1830

Original description

Test rotaliform, dorsal side flat, ventral side convexly rounded, umbilicate, periphery rounded, lobated; chambers, few (usually about four) in the last formed coil, inflated, rapidly increasing in size; sutures distinct, curved, deep, not limbate; wall granular or subspinose, very finely perforate; aperture an elongate opening extending from the umbilicus, where it is widest, almost to the peripheral margin and sometimes provided with a narrow lip.

Designation of a Neotype

Neotype

USNM PAL 781795 National Museum of Natural History, Washington, D.C., USA (Figs. 2.12.6).

Axial view (Figs. 2.12.2)

The weakly concave spiral segment is formed by the raised outer margins of chambers in the outer whorl. A very low central dome formed by earlier whorls is visible in oblique view (Fig. 2.5), as is the low height of the keel above chamber surfaces. From their acute junction with the spiral face, the convex umbilical walls of the fth (=last-formed) and f-2th (=ante-penultimate) chambers curve regularly toward the umbilicus; the axial extension of chambers in the outer whorl increases progressively, and their apices are broadly rounded. The lip of the weakly crescentic aperture is raised and encrusted near the umbilicus but declines in height. Pustules on chambers of the outer whorl are largest around the umbilicus.

Spiral view (Figs. 2.42.6)

The profile is subquadrate, with four chambers in the outer whorl that are weakly lobate, apart from the last chamber (fth). They increase in size approximately isometrically, but expansion is greatest between the f-1th and the fth, which enhances the subquadrate profile. The outer margin of the fth chamber is an elongated, weakly convex curve, but curvature increases sharply above the aperture. Sutures are very shallow. There is a low topographic ridge (keel) along the outer margin of the f-3th–f-1th chambers that continues onto the fth; it weakens and disappears above the aperture. Chamber walls are porous, but pores are covered by the keel at the outer margin. Pustules are prominent, particularly at the leading edge of each chamber and along their outer margins.

Umbilical view (Fig. 2.3, oblique view)

The umbilicus is narrow. Sutures are impressed and weakly convex in the direction of coiling. The apertural lip atrophies markedly away from the umbilicus.

Maximum dimensions (axial view)

Width = 440 μm; Height = 325 μm.

Type locality

Lomita Quarry, ‘Middle bed’; Pleistocene. Sample provided by James Kennett from a collection by Orville Bandy (Scott et al., 2015).

Diagnoses

From Truncorotalia truncatulinoides (d’Orbigny), proposed by Cushman & Bermúdez (1949) as type species; all morphogroups of Truncorotalia crassaformis are distinguished by c. 4 chambers in the outer whorl. Truncorotalia tosaensis has five chambers (±0.5; Scott et al., 1990) in the outer whorl.

Early Pliocene populations are distinguished from the closely homeomorphic Pliocene Globorotalia puncticulata (Deshayes) by their low-arched to slit-like apertures, although the height of the opening may overlap (Scott, 1980, fig. 7).

Although the aperture shape of Globorotalia juanai Bermúdez & Bolli resembles that of Truncorotalia crassaformis, it is distinguished from the Pliocene populations by its biconvex axial outline, weakly compressed periphery, and weak extension of umbilical walls in the direction of the coiling axis. Outer margins of chambers (spiral view) are regularly crescentic. Bicknell et al. (2018) showed from geometric shape analysis of samples between 4.519–5.886 Ma at DSDP Site 593 that there is a marked shift in axial shape c. 5.1 Ma, which they interpreted as marking the appearance of Truncorotalia crassaformis.

Remarks

Under ICZN (1999) a name-bearing type provides the “objective standard of reference for the application of the name it bears”; ‘objective standard of reference’ signifies that a valid definition of a taxon must be such that it includes the type specimen. While ‘objective standard’ does not imply that the architecture of the specimen is typical of a taxon’s populations, it denotes that it is a ground-truth or voucher specimen.

Comparison with the holotype

Galloway & Wissler’s (1927, fig. 12b) illustration of the holotype in axial orientation shows the spiral surface of the fth chamber rising toward its outer margin and a very low central dome formed by early whorls; these features are seen in the proposed specimen. Umbilical walls of chambers in the outer whorl progressively elevate axially, as in an isometric growth model. Properties that differ include greater curvature of the apertural opening and at the junction of spiral and umbilical walls of the proposed neotype. There is no indication of a keel in the holotype figure, but no photographic or digital imagery is available of the holotype prior to the loss of its fth chamber. A keel was visible in the damaged holotype viewed in 2008 (Scott et al., 2015, fig. 6). From that imagery and from specimens in this study, it is concluded, contrary to Scott (2019, p. 94), that a keel was developed on all chambers of the outer whorl of the holotype. The present analysis does not support my contention (Scott, 2019, p. 94) that a keel was likely absent on the final chamber of the holotype.

In spiral orientation, the holotype (Galloway & Wissler, 1927, fig. 12c) differs in its more strongly lobated outline and particularly in the shape and size of the fth chamber whose outer margin has much greater curvature. These attributes reduce its resemblance to the subquadrate shape of the proposed neotype. However, the shape of the damaged holotype (minus its fth chamber) is subquadrate (Scott et al., 2015, fig. 6).

Morphogroups and taxonomy

Variation in the morphospecies Truncorotalia crassaformis over its c. 5 Myr history is widely recognized (e.g., Parker, 1962; Lidz, 1972; Saito et al., 1981; Chaisson & Pearson, 1997; Bylinskaya, 2005; Brummer & Kucera, 2022). Bolli & Saunders (1989) called it a plexus and reviewed taxa that formally encapsulate variation in shell morphology. A limitation of this approach is that it fails to identify variation in shell form at the population level.

Analysis of the Lomita data (Figs. 1B1C) resolves two endpoints (ventroconical and biconvex) and a clump of specimens in which confidence in their partitioning is very low. The architecture of ventroconical shells (e.g., Fig. 1A, specimens 21–22 and the neotype, hereafter termed v-cones) is a four-chambered homologue, sometimes close, of the five-chambered Truncorotalia truncatulinoides. Their outlines are cone-like; the spiral surface has low relief, concave in some, while the umbilical walls taper towards the coiling axis and are strongly elevated. The periphery of the shell is angular and commonly has a weak keel. Growth of these specimens in late ontogeny is almost isometric: chambers in the outer whorl increase in size rapidly with only minor modifications to their shape. At the other pole are shells like specimens 47–48 (Fig. 1A) whose biconvex axial shape is created by the smaller dimension of the final chamber relative to its predecessor (kummerforms of Berger, 1969). Recognition of v-cone shells may be useful for research on the plexus because Quillévéré et al. (2013) found covariances in Truncorotalia truncatulinoides between the shape of its axial outline, molecular profiles, and habitat. As both taxa are deep-dwelling (Boscolo-Galazzo et al., 2022), some elements of this relation may apply to Truncorotalia crassaformis. Attention is drawn to several taxa in which v-cones are present. Recall that this classification of specimens in a common orientation is independent of their taxonomic status.

Globorotalia crassaconica Hornibrook (Figs. 2.102.12), which occurs in the Southwest Pacific subtropical gyre between c. 3.5–2.9 Ma (Crundwell, 2018), is the earliest taxon with well-developed v-cone shells. The last-formed chamber in some (Fig. 2.10) is closely homomorphic with that of Truncorotalia truncatulinoides. The spiral outline is subquadrate but the outer margin of late-formed chambers (Scott et al., 1990, fig. 69) differs from v-cones in the Lomita collection. Notably, this taxon has not been identified as the predecessor of Truncorotalia truncatulinoides (Lazarus et al., 1995).

Globorotalia crassacarina Scott, Bishop & Burt (Figs. 2.132.14), which appeared in the Southwest Pacific at c. 2.2 Ma, includes specimens that most closely resemble the proposed neotype and similar v-cone specimens in the Lomita collection (Scott, 2019, figs. 5.13–14). Its populations extend through the Quaternary.

Globorotalia hessi Bolli & Premoli Silva (Figs. 2.7–2.9) was described as a very low trochospiral shell with a flat spiral face, a strongly convex umbilical side, a roughly quadrangular outline with a faint peripheral keel on some. The authors remarked that the spiral face is often concave. Although these features suggest v-cone architecture, their SEM imagery is inconclusive. The entry of the taxon marked the base of their Pleistocene Globorotalia hessi subzone, and it was mapped in Caribbean holes to c. 80 ka (Bolli & Premoli Silva, 1973, fig. 2). Chaproniere (1991) recorded Globorotalia hessi in Coral Sea cores where it appeared in Marine Isotope Stage (MIS) 17 and extended to MIS 2 (∼700–20 ka; Berger et al., 2015). In North Atlantic cores, Bylinskaya (2005) found that Globorotalia hessi appeared at 4.5 Ma and was present into the late Quaternary. She regarded the Globorotalia hessi subzone as representing its acme.

In the equatorial North Atlantic core 234 (5.75°, –21.72°), Lidz (1972) found that ventrally vaulted, partially carinate specimens are common between MIS 8–1, whereas in core 280 (34.95°, –44.27°), similar specimens are prominent in MIS 7 and 5 but scarce and smaller in MIS 1. Although her analysis is descriptive and specimens in the imagery are inconsistently orientated, it documents a record of specimens similar to v-cones and may support her thesis that their distribution is environmentally related, specifically to temperature.

In concert, these data are consistent with the presence of v-cones in Lomita Marl, which Ponti (1990, fig. 117) showed (from molluscan amino-acid ratios) was deposited between MIS 15–11 (∼600–350 ka).

The proposed neotype is selected using an analysis of the axial shape of 54 specimens from Lomita Marl, including its type locality. The selection takes into account information about the holotype and recommendations in the ICZN. Due to the paucity of well-preserved specimens at the type locality resembling the proposed neotype, reference populations are needed to characterize Truncorotalia crassaformis as a morphospecies. The present data suggest that Globorotalia crassacarina and Globorotalia hessi may be junior synonyms. The former appeared in the lower Pleistocene, extends through the interval at which Lomita Marl was deposited (MIS 15–11, ∼600–350 ka; Ponti, 1990, fig. 117), and is present in the Holocene at DSDP Site 591A, sample 1-1-1 cm (31°35.06 S; 164°26.92 E; 2142 m water depth).

I thank James Ingle for supplying specimens, Bruce Hayward for discussions on procedure, and Ralf Schiebel and Christoph Hemleben for imagery. Appendices 1–3 can be found linked to the online version of this article.

APPENDICES

Appendix 1. Formation and locality of specimens 1–54 in Figure 1.

Appendix 2. Text file with R commands and guidance for plotting outline coordinates and executing a procrustes shape analysis of the data in axial_data_Scott_JFR2023.Rdata.

Appendix 3. The file axial_data_Scott_JFR2023.RData which contains the outline coordinates of specimens 1–54 in Figure 1 as a binary R file. It may be viewed in an R computing environment (available from https://cran.r-project.org/bin/windows/base) using the ‘load’ command.

Supplementary data