Reappraisal of three calcareous nannofossil species: Coccolithus crassus, Toweius magnicrassus, and Toweius callosus

Type material of calcareous nannofossil index species Coccolithus crassus and two geographically widespread species Toweius magnicrassus and T. callosus have been studied by both light and SEM microscopy and morphometric measurements were made. Coccolithus crassus resembles Coccolithus pelagicus but has a raised cycle of elements around the centre of the distal shield. It probably evolved from C. pelagicus. Both T. magnicrassus and T. callosus have three cycles of elements in distal view, which is a characteristic of Toweius. Toweius magnicrassus is larger than T. callosus. Differentiation of T. magnicrassus from T. callosus is possible and useful because there is generally a size gap between them in a given sample and they have different stratigraphic ranges. However, both T. callosus and T. magnicrassus appear to increase in size from high to low latitudes. Toweius callosus most probably evolved from Toweius pertusus in the latest Palaeocene and gave rise to T. magnicrassus in the early Eocene.


concepts of Toweius magnicrassus (Bukry) Romein and Coccolithus crassus was described by Bramlette and Sullivan
Toweius callosus Perch-Nielsen. Bukry (1971aBukry ( ) described C. (1961 from the Eocene Lodo Formation of California. The magnicrassus from light microscope study and the holotype first occurrence of this species is a zonal marker for the was illustrated beside C. crassus to show the different rim CPlO/CPll zonal boundary in the widely used nannofossil optics. The species name, chosen because of central area zonation of Okada and Bukry (1980). This marker could be similarities, created a false impression that T. magnicrassus is used for biostratigraphy by Bukry in virtually all lower a larger form of C. crassus, even though the original text Eocene DSDP cores that he examined (Fig. 1). However, most nannofossil workers cannot use this marker because of confusion about the species concept of C. crassus. For instance, Matter et al. (1974, pl. 4 (1985, p. 433, figs. 3.46 and 3.47, p. 504, figs. 58.12 and 58.13) figured Toweius callosus Perch-Nielsen as C. crassus. The problem of using this marker was echoed by Filewicz and Hill (1983, p. 51).
Associated with the C. crassus problem is the general confusion in the literature about the species  (Perch-Nielsen, 1971). The only light micrographs (isotype) of this species (Perch-Nielsen, 1971, pl. 61, figs. 32 and 33) unfortunately were of Reticulofenestra dictyoda (R. sarnodurovii of some authors). Besides, the size range of this species was not described. Confusion of this species with T. rnagnicrassus and C. crassus thus arose frequently in the literature. For example, Romein (1979) illustrated T. rnagnicrassus as T. callosus; Gallagher (1989, p. 47) considered T. callosus as a junior synonym of C. crassus.
Coccolithus crassus is a marker species with one of the widest geographic distributions in the Cenozoic (Fig. 1) and Toweius rnagnicrassuslT. callosus are generally abundant from low through high latitudes (see, for example, Bukry [1971a], Wei and Wise [1989], Pospichal and Wise [1990]). It is thus important to clear up the confusion about the species concepts so that these species can be used widely and consistently in biostratigraphy and in palaeoecological studies. We have re-examined type material of these species and photographed the species both in a light microscope and scanning electron microscope (SEM). Morphometric measurements for the three species were made in the type samples and in samples from different stratigraphic levels and different latitudes.

MATERIAL AND M E T H O D S
Topotype material of Coccolithus crassus is from a split of Bramlette's original sample (Lodo #68, holotype was from Lodo #71, which was not available), the type material of Toweius rnagnicrassus (DSDP 478-7-3, 104 cm) was obtained from the DSDP Core Repository at Scripps Institution of Oceanography, and the type material of Toweius callosus (KPN53) was kindly provided by K. von Salis Perch-Nielsen. Smear slides were made directly from unprocessed samples and examined with a light microscope at a magnification of about 1250X. For SEM study, samples were mounted on a cover glass glued to a specimen stub and coated with a thin film of gold-platinum alloy in a vacuum coater.
For morphometric studies, at least 30 (typically >50) specimens encountered along random traverses of each smear slide were measured with a ruler on a Panasonic monitor screen connected to a Panasonic video camera mounted on a Zeiss Photomicroscope 111. The magnification we achieved in this study is 6600X, that is, one cm on the screen corresponds to about 1.5pm for a fossil specimen. This enables a size resolution better than 0.3pm, which is sufficient for this study. most distinctive feature of C. crassus is a raised cycle of elements around the centre of the distal shield (Pl. 1, fig. 1; see also Wei [in press]). This cycle can also be seen easily in a light microscope by focusing up and down through the specimen. Although the elements of the raised cycle appear to be slightly irregular, they are not likely to be the product of overgrowth since they are consistently present in the topotype material, which contains generally well-preserved nannofossils, and in age-equivalent material from DSDP Sites 47 (Pl. 1, fig. 1) and 528. Without the raised cycle, the species would look very similar to Coccolithus pelagicus (Wallich) Schiller. On the other hand, C. crassus differs significantly from species of Toweius in that it shows only two cycles of elements in distal view whereas Toweius shows three cycles of elements. C. crassus thus should not be transferred to the genus Toweius as proposed by Perch-Nielsen (1984) but should remain in the genus Coccolithus. The reasons for the use of Coccolithus rather than Ericsonia have been discussed by Wise (1983).

Light and SEM micrographs of
Coccolithus crassus can be differentiated from C. pelagicus by the following features: (1) in phase-contrast light, the distal shield of C. crassus is less dark than that of C. pelagicus; (2) the area around the central opening is very bright; (3) the distal and central margins of C. crassus show slight irregularities mimicking overgrowth; (4) in polarized light, an orange line is present close to the central opening of C. crassus, and virtually the entire placolith appears to be birefringent (see P1. 1. fig. 7).
Measurements on C. crassus (Fig. 2) show that the length/ width ratio (mean=1:0.83) is relatively constant for specimens of different sizes and the species is elliptical. The size range is commonly 8-12 pm. This size range is in good agreement with the holotype size (9.5pm), although Bramlette and Sullivan (1961) gave a size range of 10-13pm for the species.
Toweius rnagnicrassus was described by Bukry (1971a) from the lower Eocene of DSDP Site 47 in the northwestern Pacific. Both holotype and isotypes are light micrographs. Published SEM micrographs of Toweius magnicrassus are rare, the two in Wei (1992, pl. 1, figs. 1 and 2) are probably the only ones available. Here we present SEM and light micrographs of T. magnicrassus from the type material (Pl. 1, figs. 4, 5, and 8-11). Romein (1979) transferred Coccolithus niagnicrassus to the genus Toweius. However, as pointed out by Perch-Nielsen (1985, p. 505), the two specimens he illustrated as T. magnicrassus (pl. 4, figs. 2 and 3 ) are 6.5pm and 7.2pm long, respectively, and therefore too small to be T. magnicrassus. They are actually specimens of T. callosus. Thus it was not clear whether C. magnicrassus belonged to Toweius and Perch-Nielsen (1985) attached a question mark to Romein's new combination. It is now clear that the question mark should be removed because T. magnicrassus clearly shows three cycles of elements in distal view (see P1. 1, figs. 4 and 5), which is characteristic of the genus Toweius. It is also clear that T. magnicrassus is not a larger form of C. crassus because the two species do not even belong to the same genus.  1, figs. 2, 3, 12, and 13). As T. magnicrassus, T. callosus clearly shows three cycles of elements in distal view, and in fact, the structure of T. callosus is virtually the same as that of T. magnicvassus. However, there is generally a size gap between the two species and they have different stratigraphic ranges. It is thus useful to distinguish the two species. Perch-Nielsen (1985, p. 505) remarked that "T. callosus is difficult to distinguish from Neogene species of Reticulofenestra or from early forms of R. dictyoda with the L M (light microscope). It is now clear that the birefringence pattern of Toweius callosus is very different from that of Reticulofenestra because the outer rim of T. callosus birefringes very weakly (Pl. 1, figs. 12 and 13) whereas the entire placolith of Reticulofenestra birefringes strongly.
In order to investigate the size patterns of T. magnicrassus-T. callosus through time, we made morphometric measurements on eight samples from different stratigraphic levels at DSDP Site 47 (Fig. 3). Previous nannofossil biostratigraphy for this site was provided by Bukry (1971b). Here we redated this interval using the nannofossil zonation of Okada and Bukry (1980) to achieve higher biostratigraphic resolution. Samples 47B-7-2, 30 cm through 47B-7-3, 104 cm contain Discoaster sublodoensis, Discoaster kuepperi, and Coccolithus crassus among other Eocene species but no Nannotetrina fulgens, and can be assigned to Zone CP12. Discoaster sublodoensis and Tribrachiatus orthostylus are not present in Sample 47B-4, 30 cm whereas Discoaster lodoensis, Discoaster kuepperi and C. crassus are abundant, and this r 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 7 7.5 8 8.5 9 9.5 10 10.5 1 1 11.5 Size (pm) Size (pm) Length (pm) Length (pm) orthos$us but no C. crassus and this interval is dated as Zone CPIO. Sample 47B-8-3, 30 cm contains Discoaster diastypus but no Discoaster lodoensis and can be assigned to Zone CP9. Samples 478-7-2, 30 cm through 478-7-5, 30 cm contain T. magnicrassus but no T. callosus (Fig. 3). It is clear from these samples, including the type sample of T. magnicrassus (Sample 47B-7-3, 104 cm), that the size of T. magnicrassus generally ranges from 9 to 15pm. The 16-20pm size range for this species in the original description of Bukry (1971a) thus should be revised. Measurement of the holotype (Bukry, 1971a, pl. 2, figs. 1 and 2) using the magnification stated results in a size of 15pm. The length vs. width diagrams (Fig. 3) show that there is a relatively large range in the length vs. width ratio (mean=1:0.75) and that a few specimens are circular whereas most specimens are elliptical to subelliptical.
Toweius callosus is abundant in Samples 47B-8-1, 30 cm through 47B-8-3, 30 cm (Fig. 3). A few specimens of T. magnicrassus have been recorded in Sample 478-8-1, 30 cm, which are 29 pm in size. It is noticeable from Figure 3 that there are generally more circular or virtually circular specimens in T. callosus than in T. magnicrassus. In fact, the holotype of T. callosus is close to circular.
The morphometric measurements of the T. mugnicrassus/T. callosus group in the type material of T. callosus (KPN53) are shown in Fig. 4. T. callosus is very abundant in this sample whereas T. rnagnicrassus is more than an order of magnitude less abundant, so we measured these two species separately to reveal the size distribution of both species. Separation of the two species based on size is possible because there seems to be a size gap between them, with 6.5-7.5pm sized specimens being absent. T. callosus is 4 to 6pm long and T. magnicrassus 8 to 11pm (Fig. 4). This is consistent with most of the illustrations of Perch-Nielsen (1971), where the holotype of T. callosus is 5.4pm long (Perch-Nielsen, 1971, pl. 17, fig. 5), and the isotypes are 5.1pm (Perch-Nielsen, 1971, pl. 17, figs. 3 and 5). However, one of her isotypes of T. callosus (Perch-Nielsen, 1971, pl. 18, fig. 5) is not T. callosus because that specimen is about 8.0pm long and it belongs to T. magnicrassus.
T. magnicrassus and T. callosus are both smaller in highlatitude (56"N) Sample KPN53 than in low-latitude samples from DSDP Site 47 (DSDP Site 47 lay in low latitudes in the early Eocene [Prince et al., 19801). In order to better examine latitudinal variation in T. magnicrassuslT. callosus, we also measured the size of T. magnicrassus and T. callosus in two samples from mid-latitude Site 605 (39"N). These results are compared with those from the high and low latitudes in Fig.  5. There appears to be a size increase for both T. callosus and T. magnicrassus from high to low latitudes. The upper size limit of T. callosus in the low-latitude samples overlaps with the lower size limit of T. magnicrassus in the high-latitude sample, although these two species generally do not overlap in size at individual sites. Consequently, the definition of the size range of the two species should take into account this size shift through latitude. This sizelatitude relationship may have significant implications in palaeoecologic studies. However, a detailed investigation of this is beyond the scope of the present paper.
The evolution of the three species investigated is either not clear or incorrectly stated in the literature, partly because of the confusion of the species concepts. For instance, Romein (1979, p. 72, fig. 38) suggested that T. magnicrassus evolved directly from Toweius pertusus. He did not discuss the evolution of T. callosus or C. crassus. Gallagher (1989, p. 49, fig. 3.4) indicated that T. pertusus evolved to T. crassus (his T. callosus, [Gallagher, 1989, p. 47 ranges are different. There is, however, a general size increase for both T. callosus and T. magnicrassus from high to low latitudes. Toweius callosus is usually 4 to 6pm long at high latitudes, 4 to 8pm at mid latitudes, and 6 to 8.5pm at low latitudes. Toweius rnagnicrassus is generally 8 to llpm long at high latitudes, 9 to 13 mm at mid latitudes, and 9.5 to 14.5pm at low latiwdes. Toweius callosus most probably evolved from Toweius pertusus in the latest Palaeocene (Zone CP8) and gave rise to T. rnagnicrassus in the early Eocene.  fig. 8.