Cassidulina teretis Tappan and Cassidulina neoteretis new species (Foraminifera): stratigraphic markers for deep sea and outer shelf areas

Fossil versus Recent specimens allocated to Cassidulina teretis Tappan display slight differences in both ecological distribution and morphology. This has led to a re-examination of specimens from the North Atlantic region. The study was mainly based on scanning electron microscopy. It resulted in the division of the plexus into two species with different stratigraphical and partly different environmental distributions: Cassidulina teretis Tappan and Cassidulina neoteretis n. sp. A distinction between C. teretis and C. neoteretis provides a new biostratigraphical marker and may also prove useful in ecological studies. C. teretis had its first occurrence during the Middle to Upper Miocene and its last well-documented occurrence shortly after the palaeomagnetic Brunhes/Matuyama boundary in the Norwegian Sea, whilst it apparently disappeared from the North Atlantic as early as a little after the Gauss/Matuyama boundary. C. neoteretis presumably evolved from C. teretis between about 2.0 and 2.3 Ma in the northern North Atlantic and migrated northward inhabiting the Norwegian Sea as C teretis became extinct here at about 0.7 Ma. All Recent specimens belong to C. neoteretis. Specimens of C. teretis have been documented from both arctic and boreal regions in inner shelf to bathyal environments (between about 50 and 2000 m water depth), whereas the Recent distribution of C. neoteretis is slightly more limited: arctic and cold boreal regions at water depths between 150 and 3000 m, most commonly between 1000 and 1500 m.


INTRODUCTION
The species Cassidulina teretis was first described by Tappan (1951) from the Pleistocene of the Gubik Formation, Alaska. Since then the name has been used for specimens from various localities spanning Miocene to Recent times.
Fossil specimens assigned to C. teretis are documented from both inner shelf to bathyal environments ranging from arctic to boreal regions. The Recent specimens, which hitherto have been referred to as C. teretis on the other hand have a more restricted distribution ranging from outer shelf to bathyal environments in arctic to cold boreal regions. Fossil and Recent specimens also appeared to show small differences in morphology, a fact that has been commented on previously by Mackensen & Hald (1988). Because of these environmental and morphological differences a re-examination of the taxon was carried out. The purpose of this paper is to describe this morphological difference and the stratigraphical and ecological implications.

MATERIAL
Authors of several previous studies graciously provided specimens for this study (see Acknowledgements below). Their Recent material is from the central Arctic Ocean, Prince Gustav Adolf Sea, Scoresby Sund, the East Greenland continental margin, the Labrador Sea, the Irminger Basin, the Norwegian Sea northeast of Iceland, the Norwegian continental margin, the Barents Sea, and the Weddell Sea (Fig. 1). Fossil specimens are from Baffin Island (Clyde Foreland and Qivituq), East and Northeast Greenland (Lodin Elv and Kap Kobenhavn), Iceland (Tjornes), the northern North Atlantic, the North Sea and the Norwegian Sea (Fig. 1).
About 2000 specimens of Cassidulina have been examined under a light microscope, and approximately 400 of these were later studied usilig a scanning electron microscope. Systematics follow Nomura (1983a,b) and Loeblich & Tappan (1988).

NEW PLEISTOCENE BIOSTRATIGRAPHICAL MARKERS
Specimens assigned to Cassidulina teretis Tappan (1951) have been recorded from the Upper to Middle Miocene to Recent. A comparison of specimens from this interval with the holotype of C. teretis (USNM no. 560409, Smithsonian Institution, Washington D.C.) has revealed that a small but general change in the morphology of the specimens has occurred within the stratigraphical record. Specimens that are found prior to this change do not differ from the holotype, whereas the apertural flaps of younger specimens are always much smoother than that of the type specimen of C. teretis (see Systematics below). This difference is small and would not be of any consequence, were it not for the possibility of using this morphological change as a Pleistocene biostratigraphical marker event.
To prevent confusion in distinguishing between the older and younger types a new name is introduced for the latter. The younger type with the smooth apertural flap is given the name Cassidulina neoteretis n. sp. All Recent specimens recorded (see below) belong to C. neoteretis. Cassidulina teretis is probably extinct.   In order to establish exactly when the C. teretis LAD and the C. neoteretis FAD took place a number of records have been examined. All specimens from examined Neogene records prior to about 2.3 Ma (Figs 1, 2) have proven to be of true C. teretis (Feyling-Hanssen, 1980, 1985, 1990b: Feyling-Hanssen et al., 1983: Murray, 1987Schnitker, 1984: Diester-Haass & Schnitker, 1989: Jansen et al., 1990: Knudsen & Asbjornsdbttir, 1991. True C. teretis have been recorded until slightly above the palaeomagnetic Brunhes/Matuyama boundary in the North Sea and in ODP Hole 644A from the Voring Plateau in the Norwegian Sea (Fig. 2) (Sejrup et al., 1987, Jansen et al., 1990. These are the youngest records of C. teretis that have been dated with certainty. The species has been described from possibly younger sediments of the Ymer Formation, northeastern Greenland, but the dating of this deposit is uncertain (Hjort & Feyling-Hanssen, 1987;Feyling-Hanssen, 1990a).
The oldest record of C. neoteretis from the Norwegian Sea is also just above the Bruhnes/Matuyama boundary. This record is from ODP Hole 642C (Osterman & Qvale, 1989) which is located close to Hole 644A (Fig. l), and the dating is obtained through correlation with Hole 642B from the same site (see Henrich, 1989). None of these records show the transition from C. teretis to C. neoteretis, but it is likely that in the Norwegian Sea the LAD of C. teretis and first occurrence of C. neoteretis both took place at approximately 0.7 Ma.
In the North Atlantic (DSDP Hole 552A), on the other hand, C. neoteretis already replaced C. teretis about 2 million years ago ( Fig. 2) (data from Shackleton et al., 1984: Murray, 1987Diester-Haass & Schnitker. 1989) (Fig. 2). This is evidence of diachronism in the C. teretis LAD and the C. neoteretis FAD across the North Atlantic region (Figs Unfortunately, C. teretislneoteretis were absent or near absent from Hole 552A during the time interval from about a little after the Gauss/Matuyama boundary (Core 9, section 2; about 2.3Ma) to just prior to the palaeomagnetic Reunion event (Core 8, section 2: about 2.0 Ma), making an exact placement of the morphologic change from C. teretis to C. neoteretis impossible. At the same time, the lack of C. teretis after about 2.3 Ma implies that the disappearance of C. teretis on Baffin Island and East Greenland close to the Gauss/Matuyama boundary ( Fig. 2) (Feyling-Hanssen, 1980, 1985Feyling-Hanssen et al., 1983) may actually have been a stratigraphic event linked to the extinction of C. teretis in the North Atlantic.
The profound similarity between C. teretis and C. neoteretis indicates that the latter species is probably an evolutionary descendant of the first. Thus, C. neoteretis possibly evolved from C. teretis in the North Atlantic between about 2.0 and 2.3Ma. It gradually migrated northwards and replaced C. teretis in the Norwegian Sea when this species became extinct at about 0.7 Ma BP. (Fig.  2). It is notable that C. neoteretis did not migrate to all areas where C. teretis had lived. An example is the North Sea where C. teretis is a common species in the early Pleistocene. After the LAD of C. teretis at about 0.7Ma the plexus disappeared and C. neoteretis does not live in the North Sea today. The only record of C. neoteretis from this region is from Boring 3008 from the Gullfaks Field (K.L. Knudsen & R.W. Feyling-Hanssen, unpubl. data), where C. neoteretis is found in a single sample from the Upper Pleistocene (Fig.  2). This single occurrence should be seen as of only local importance.
The occurrence of C. neoteretis in the Weddell Sea of Antarctica (Mackensen & Douglas, 1989;Mackensen et al., 1990) may be the result of a southward migration during glacial periods.

Cassidulinu teretis Tappan and Cassidulinu neoteretis new species
In the eastern North Atlantic, the southern limit of C. neoteretis' main distribution coincides with the Iceland-Faeroe Ridge (Fig. 1) (see Jansen & BjGrklund, 1985;Mackensen et al., 1985;Mackensen, 1987;K. B. Christensen, unpublished data). It is found on the Rockall Plateau but occurs otherwise only sporadically south of this limit (J. F. Weston in Murray, 1987;Diester-Haass & Schnitker, 1989; own unpublished data). In the western North Atlantic region, its distribution extends south through the Irminger Basin, west of and on top of the Reykjanes Ridge, to the Labrador Sea, where the southernmost record is northeast of Newfoundland ( Fig. 1) (Cole, 1981;Schafer & Cole, 1982;G. Bilodeau, pers. comm., 1992).
Cassidulina neoteretis is infaunal (Jansen et a / . , 1990;Murray, 1991) and is often associated with fine-grained, organic rich, terrigeneous mud at bottom water temperatures of about -1°C and salinities of about 34.9% (Mackensen et al., 1985;Mackensen & Hald, 1988). In the Labrador Sea, it is recorded at bottom water temperatures as high as 3.5"C (G. Bilodeau, pers. comm., 1992). The distribution of C. neoteretis is generally dependent on the distribution of water masses (Mackensen, 1987), but the species is sometimes restricted to a specific depth interval of a particular water mass. In the Labrador Sea the species occurs in the Denmark Strait Overflow Water, but only in the upper part (above 3000 m) of the underlying Northeast Atlantic Deep Water (G. Bilodeau, pers. comm., 1992). An accurate record of the distribution of C. neoteretis is hampered by its frequent confusion with species such as C. laeuigata s.l., Paracassidulina neocarinata, and Islandiella helenae Feyling-Hanssen & Buzas.
Fossil records of C. neoteretis exist from both relatively shallow and deep water palaeoenvironments. The only North Sea record of the species is from a single sample of Upper Pleistocene sediments (Fig. 2) presently at a water depth of 140 m; the depositional depth did not exceed 100 m (K. L. Knudsen & R. W. Feyling-Hanssen, unpublished data). Its abundant presence in this sample may be due to temporary changes in the current system allowing Norwegian Sea water masses to enter the North Sea.
In the Andfjorden and Malangsdjupet troughs at the Norwegian continental shelf (present water depth: 200-400m), C. neoteretis was a common species during most of the last deglaciation, when the area was characterized by cold, ice-free surface water. It was replaced by C. laeuigata s.1. in the Holocene after the establishment of the Norwegian Current with its warm, saline surface water (Mackensen & Hald, 1988;Jansen & Bjorklund, 1985). A similar distribution is recorded in a boring at 1125 m water depth south of the Faeroe Islands (Boring 1007, K. B. Christensen, unpublished data). C. neoteretis is also mainly associated with glacial periods at DSDP Hole 552A in the North Atlantic (Diester-Haass & Schnitker, 1989).
Contrary to this, records from deep water in the Norwegian Sea show an immigration of C. neoteretis to the Norwegian Sea during the last deglaciation and a temporary decrease in its frequency during the Younger Dryas cold period (Jansen & Bjorklund, 1985;Jansen & Erlenkeuser, 1985).
Cassidulina neoteretis is, thus, an indicator of cold water masses in the northern North Atlantic and on the Norwegian shelf, while it may characterize ameliorated conditions in deep water areas of the Norwegian Sea.

Cassidulina teretis
Cassidulina teretis is equally abundant in the fossil record of deep sea and shelf areas with palaeo-water depths of 50-2000 m Schnitker, 1984, Diester-Haass & Schnitker, 1989Feyling-Hanssen, 1985, 1990aKnudsen & AsbjornsdBttir, 1991). Cassidulina teretis had its first occurrence in the Norwegian Sea and the North Sea during the warm climate of the Middle to Upper Miocene (Jansen  et a / . , 1990; Laursen et a / . , 1992). The species characterizes pre-glacial Pliocene deposits from the North Sea, Canada and Greenland and decreases in frequency at the onset of Northern Hemisphere glaciations (e.g. Feyling-Hanssen, 1985;Knudsen & AsbjornsdBttir, 1991;Seidenkrantz, 1992). In the North Atlantic the species presumably disappeared near the onset of the Northern Hemisphere glaciations at about 2.4 Ma. In northeastern Greenland, it is recorded only in interglacial deposits (Feyling-Hanssen, 1990a).
These records show that C. teretis may be considered a warm-water indicator in polar and possibly also in boreal regions. Hence, C. teretis had a habitat relatively similar to that of C. neoteretis, but it more often inhabited shallow and relatively warm water deposits. Thus, C. teretis seems to have had a wider ecological range than C. neoteretis. Description. Test lenticular, biconvex with an acute, slightly undulating, peripheral margin. Umbilical boss of milky. semitranslucent shell material on each side. Eight to ten chambers (frequently 10) in the final whorl, biserially arranged in 4 to 5 alternating pairs, each chamber appearing large, rounded rhomboid to ovate and reaching to the umbilical boss on one side of the test, and small and subtriangular on the other side. Sutures distinct, thickened, but not limbate, slightly depressed and outlining the chamber. Wall calcareous, hyaline or opaque, and optically granular. Surface smooth with relatively small, rounded, pores evenly distributed on the chamber walls but with no pores on the umbilical boss or along the sutures. Aperture an elongate, narrow slit extending from the base of the final chamber in a crescent paralleling the outer margin of the chamber, reaching 213 to 314 the distance from the base of the chamber to the peripheral keel. A subtriangular apertural plate with a smooth edge, formed by the infolded chamber wall, lies along the inner margin and partly covers the aperture (a type H I aperture with a slight tendency to a type H2 aperture, according to definition by Nomura (1983a) Variation. There is very little variation within this species. The edge of the apertural plate, however, varies between being totally smooth (normal) and having very small, rounded serrata as in the specimen P1. 3, fig. 5. Furthermore, the shape of the apertural plate varies between triangular (normal) and crescent-shaped. The periphery may have a keel or carina, which may have small ruffles (see P1. 2, fig. 9). Afinities. Cassidulina neoteretis shows close resemblance to Cussidulina tereiis Tappan, 1951, but differs in having a smooth, normally somewhat broader apertural plate with a more distinct angle at the marginal end giving a subtriangular shape, while C. ieretis has a crescent-shaped, serrate, apertural plate (see PI. 4, figs 1-5). Furthermore, C. neoieretis is normally smaller in size. The differences are small, but are in well-preserved specimens detectable in both light microscope and scanning electron microscope.
Cassidulina neoteretis also shows some affinity to Paracassidulina neocarinuta (Thalmann), but this species has more ovate and curved chambers which are generally fewer in numbers (4 pairs). Furthermore, the aperture of P. neocarinata is a long, narrow slit covered by a very thin, smooth lip (type G aperture, according to definition by Nomura, 1983a), and both the umbilical boss and the pores are very small (see emended description by Nomura (1983b) and PI. 1,P1. 4,fig. 7;PI. 5 ,figs 6,9,10).

Remarks.
A thorough description of this species was made by Tappan (1951). Only an emendation of the apertural description will be added here: Tappan (1951) described the aperture as 'elongate, extending from the base of the final chamber in a crescent paralleling the anterior margin of the chamber, reaching nearly three-fourths the distance from the base of the chamber to the peripheral keel'. The aperture is partly covered by a narrow, serrate plate formed by the infolded chamber wall (type HI aperture, according to definition by Nomura (1983a)). The serration along the edge of the apertural plate is not dissimilar, but 105, PI. 1, figs 6-9, 11-13.

Explanation of Plate 2
Scanning electron microscope photographs . Figs 1-14 The serration is most prominent in the best preserved specimens. This documents that the serration is a primary character of the species and is not due to post mortem dissolution. The size of the specimens measured in this study varies with the greatest diameter between 250 and 550pm, the bulk of the specimens, however, having a greatest diameter of 400-450 p m .
In studies from some localities from Greenland and the North Sea a somewhat smaller variety of Cassidulina teretis has been separated from C. teretis as Cassidulina cf. teretis (Feyling-Hanssen et al., 1983;Funder et al., 1985;Seidenkrantz, 1992). Cassidulina cf. teretis differs from the typical C. teretis by its often smaller size and smaller umbilical boss. A thorough examination of the two types have, however, shown transitional forms between them. Furthermore, there is no difference in the shape of the aperture. This leads to the conclusion that C. cf. teretis is a morphological variation of C. teretis.
Examination of paratypes of C. teretis have shown that several of these belong t o the genus Islandiella ( I . infrata and I. helenae) (USNM no. 560413 and part of no. 560413; see also Feyling-Hanssen & Buzas, 1976).

SUMMARY
Specimens which until now have been assigned to Cassidulina teretis belong t o two different species, Cassidulina teretis Tappan and Cassidulina neoteretis n. sp. The most apparent difference between these is the shape of the apertural flap, but the latter species is also generally smaller in size. The differences are small, but are detectable in well-preserved specimens in both light microscope and scanning electron microscope.
Cassidulina neoteretis presumably evolved from C. teretis in the North Atlantic between about 2.0 and 2.3Ma and migrated northwards at a later time. The L A D of C. teretis and the first occurrence of C. neoteretis in the Norwegian Sea was during the Middle Pleistocene, just above the palaeomagnetic Brunhes/Matuyama boundary (Fig. 2 ) . The distinction between C. teretis and C. neoteretis thereby allows the establishment of a new biostratigraphical marker.
Cassidulina teretis lived in both deep-sea and neritic palaeoenvironments of boreal to subarctic areas, whereas C. neoteretis is mainly a deep-sea species found in arctic and antarctic to subarctic water masses.