Dinoflagellate cysts and acritarchs from the Oligocene–Lower Miocene interval of the Alma-1X well, Danish North Sea

This palynological study of cuttings samples from the Lark Formation in the Danish North Sea well Alma-1X documents for the first time in the public domain the succession of last occurrences of dinoflagellate cysts and acritarchs in the Oligocene–Lower Miocene interval of the Central North Sea. The distribution of dinoflagellates and acritarchs in the well demonstrates the potential for the development of a detailed subdivision of the Oligocene–Lower Miocene in the Central North Sea, based on the first downhole occurrences of key taxa. Five regional intra-Lark Formation seismic and petrophysical log markers can be dated with precision using dinoflagellate biostratigraphy. Four new species and one new subspecies of dinoflagellates are described from the study interval: Amphorosphaeridium? almae sp. nov., Filisphaera pachyderma sp. nov., Pentadinium corium sp. nov., Spiniferites pseudofurcatus verrucosus ssp. nov. and Thalassiphora rota sp. nov. Pseudospiniferites manumii Lund, 2002 is emended and transferred to the genus Spiniferites.


INTRODUCTION
Recent oil discoveries within Palaeogene sandstones in the Danish North Sea sector have sparked interest in the Palaeogene stratigraphy of this area. The present study is a part of a larger stratigraphical project on the Palaeogene sediments of the North Sea carried out by the Geological Survey of Denmark and Greenland (GEUS). The aim of the project is to improve understanding of the distribution and geometry of the Palaeogene sediment package, in particular its sandstone units, and to revise the lithostratigraphical subdivision of the siliciclastic succession above the Upper Cretaceous-Danian chalks. The project integrates seismic and log interpretations with an improved biostratigraphy from new studies of the organicwalled dinoflagellate cyst (hereafter dinoflagellate) and acritarch succession.
This study documents for the first time in the public domain the dinoflagellate and acritarch assemblage in the Oligocene through Lower Miocene interval from a Central North Sea well. Previously published stratigraphical and taxonomic studies that document the phytoplankton succession from the Oligocene-Lower Miocene interval in the greater North Sea Basin, include Gerlach (1961), Benedek (1972), Liengjarern et al. (1980), Piasecki (1980), De Coninck (1986, 1999, Heilmann-Clausen & Costa (1989), Köthe (1990), Strauss & Lund (1992), Powell (1992), Pross (1997), Dybkjaer & Rasmussen (2000) and Strauss et al. (2001). However, none of these studies incorporated borehole data from the present North Sea which is the aim of this contribution.
The onshore Palaeogene succession at the margins of the basin has large stratigraphical gaps caused by periods of nondeposition and/or erosion (Michelsen et al., 1998). This is particularly true for post-Eocene strata and probably reflects the effects of glacio-eustatic sea-level changes as a response to climatic changes (Zachos et al., 2001). In order to document a more comprehensive picture of the Oligocene-Lower Miocene dinoflagellate and acritarch assemblages, it is necessary to incorporate data from the central and deeper part of the North Sea Basin where the stratigraphy is more complete. The Alma-1X well is located in such a setting -in the southern part of the Danish Central Graben (UTM 31 E 639612 m N 6150851 m, Fig. 1). Although not entirely free of unconformities, Alma 1X provides a more complete picture of the phytoplankton assemblages of the Oligocene and Lower Miocene than any other onshore outcrop or well provides.
The majority of published zonation schemes covering the study interval in NW Europe are based on onshore exposures or cores from onshore boreholes and focus on the stratigraphical lowest occurrences (LO) of dinoflagellates. Only a few North Sea wells have cored sections in the Oligocene or Lower Miocene interval; almost all well material is ditch cuttings samples, which are notoriously subject to downhole contamination. Therefore, dating these wells uses the first downhole occurrence (FDO) of taxa and the zonation schemes based on LO of taxa are of limited value. Major zonation schemes covering the Oligocene and Lower Miocene interval (partly or entirely) of the North Sea area (Costa & Manum, 1988;Heilmann-Clausen & Costa, 1989;Köthe, 1990;Gradstein et al., 1992;Powell, 1992;Gradstein & Bäckström, 1996) are not as detailed as those covering the Paleocene and Eocene (e.g. Mudge & Bujak, 1996). Therefore, a second aim of the present work is to explore the possibilities for establishing a more detailed subdivision of the Oligocene and Lower Miocene based on FDOs of key dinoflagellates and acritarchs.
In the interval 6110-5450# (1862.3-1661.2 m) all the cuttings samples were studied, hence there is an uphole error of 30# (9.1 m) for any event occurring in that interval. In the interval 5450-4550# (1661.2-1386.8 m) every second cuttings sample was studied, resulting in an uphole error interval of 60# (18.3 m).
The samples were processed following standard palynological preparation procedures (Batten, 1999); residues were sieved at 11 µm. A brief oxidation (2 min in 36% nitric acid) was carried out on all residues before they were separated using zinc cloride and mounted in glycerine jelly on microscope slides. Type mounts were sealed airtight with nail varnish.
The stratigraphical distributions of dinoflagellate and acritarch taxa encountered are shown in Figure 2. All taxa are listed alphabetically and referenced in a table available online at http://www.geolsoc.org.uk/SUP18218. A hard copy can be obtained from the Society library. The taxa are illustrated in Plates 1-19. LVR numbers refer to file numbers in the fossil image database of GEUS which contains all relevant curatorial details on each specimen. A SM number further identifies single specimen mounts. Types and paratypes are kept in the type collection of the Geological Museum of Copenhagen and have been given MGUH registration numbers in that collection. Figured specimens that are not type material are stored in the palynological slide collection of GEUS.
As Figure 2 indicates, reworked specimens occur throughout the well section, particularly in the upper Lark Formation. As a rule of thumb, isolated occurrences above a consistent occurrence interval in Figure 2 were interpreted as reworking unless the taxon in question is rare. In the latter case, isolated tops were accepted as in situ occurrences. However, isolated tops were not accepted for any of the key FDO events shown in Figure 2.
Dinoflagellate taxa mentioned in the text, figures and in the compiled taxa list (available online at http:// www.geolsoc.org.uk/SUP18218) and in Table 1 are cited and referenced in Williams et al. (1998) or herein. Suprageneric classification follows Fensome et al. (1993). Acritarch taxa are referenced in Fensome et al. (1990).

LITHOSTRATIGRAPHY
The lithostratigraphical subdivision of the Oligocene and Miocene follows that of Knox & Holloway (1992). All samples studied, except the lowermost, belong to the Lark Formation. The lowermost sample is from the Horda Formation, identified on log signature and cuttings sample lithology (see below).
The studied section of Alma-1X is intersected by five surfaces that can be traced on seismic sections over large areas of the North Sea (Fig. 3). These surfaces can also be picked on the gamma log from the well (Fig. 4). Three of the surfaces mark the base of prograding wedge complexes and two are discontinuity surfaces.
The boundary between the Horda and Lark formations is sharp and unconformable and characterized by a change from dark, greenish-grey to greyish-green, fissile mudstones below the boundary to light-coloured, greenish-grey, in some wells dark brown, non-fissile mudstones above the boundary. On the gamma log, this change is accompanied by an uphole increase to a consistently higher gamma response level in the Lark Formation at 6079.4# (1853.0 m) (Fig. 4). On the seismic section a conspicuous reflector (Top Horda, Fig. 3) marks the boundary.
Above the Top Horda reflector, a parallel seismic reflection pattern dominates the Lark Formation up to an Upper Oligocene unconformity (UOU Marker,Figs 3,4). In the lower part of this interval the Lark 4 Marker is indicated by a prominent seismic reflector and a spike on the gamma log followed by an uphole decrease in gamma readings. Above the UOU Marker the general seismic architecture of the basin is characterized by large clinoforms separated by onlapping sediment wedges, and the lithology changes to brown to beigebrown mudstones. Two seismic reflectors, the Upper Lark Marker (ULM) and the Lark 18 Marker (Figs 3, 4), above the UOU Marker separate major prograding sediment packages. The Lark 18 Marker, located at the top of the studied interval, separates dominantly dark, beige-brown mudstones below from dark brown mudstones above.

CHRONOSTRATIGRAPHY
Following well completion a chronostratigraphical subdivision of the Palaeogene and Neogene section was established based on microfossils (unpublished data from Robertson Group plc.). Key microfossil events recorded in that study have been used herein for correlation with the agglutinating, benthic and planktonic microfossil microfossil zones (NSA, NSB and NSP) of King (1989) established for the North Sea (Fig. 4). King attempted to calibrate his zonal markers with the standard chronostratigraphic scale of Berggren et al. (1985a, b). However, only a few first-order correlations were possible; most of the calibrations were made using dinoflagellates, planktonic foraminifera and nannoplankton from onshore sections in the North Sea Basin (King, 1989); the correlation of the Lower Miocene is particularly uncertain (King, 1989, p. 446). Calcareous microfossils can be very rare or absent in thick intervals of the Palaeogene in the North Sea (including the interval studied herein).
In the absence of a good, self-contained microfossil biostratigraphy, the chronostratigraphical breakdown of Alma-1X relies heavily on a dinoflagellate biostratigraphy based on direct or indirect correlation with onshore type sections and/or North Sea Oligocene-Lower Miocene dinoflagellates and acritarchs boundary type sections. Whenever possible, key microfossil datums have been used to support the biostratigraphy. The criteria used herein for stage and series divisions are listed below:

Priabonian-Rupelian (Eocene-Oligocene) boundary
At its Global Stratigraphical Section and Point (GSSP; Massignano section, central Italy), the Eocene-Oligocene boundary is bracketed by the highest occurrence (HO) of the dinoflagellates Cordosphaeridium funiculatum (above the boundary) and Heteraulacacysta porosa and Rhombodinium porosum (below the boundary) (Brinkhuis & Biffi, 1993). The two latter events are located in the upper part of the Eocene Series (Brinkhuis & Biffi, 1993;Vandenberghe et al., 2003). The HO of the dinoflagellate Areosphaeridium diktyoplokum is used commonly by North Sea biostratigraphers as evidence for penetration to the Eocene; however, this event occurs above the Eocene-Oligocene boundary in the GSSP type section, but at the top of the Priabonian Stage in the Priabonian type section (Brinkhuis, 1994;Brinkhuis & Visscher, 1995) thereby creating a Priabonian-Rupelian boundary problem (see Brinkhuis & Visscher, 1995). Other Eocene-Oligocene boundary sections in the type region were studied by Wilpshaar et al. (1996); however, additional key dinoflagellate events bracketing the boundary in these sections were either not encountered in the North Sea

Rupelian-Chattian (Lower-Upper Oligocene) boundary
The principal criterion for this boundary has not, as yet, been decided by the Subcommission on Palaeogene Stratigraphy. Herein, the Rupelian-Chattian stage boundary is placed at the HO of the benthic foraminifera Rotaliatina bulimoides, a conspicuous event used by most North Sea biostratigraphers as the boundary marker. The HO of R. bulimoides marks the top of the NSB7 Zone of King (1983) and the NSR7 Zone of Gradstein et al. (1994). The HO of R. bulimoides is at 29 Ma in the North Sea according to Gradstein & Bäckström (1996); slightly older than the 28.5 Ma age for the Rupelian-Chattian Stage boundary quoted by Hardenbol et al. (1998). In many North Sea wells studied by the present author, including Alma-1X, the FDO of R. bulimoides coincides approximately with the FDO of the dinoflagellate Rhombodinium draco.

Chattian-Aquitanian (Oligocene-Miocene) boundary
This boundary is bracketed by a number of HOs at its type section (Lemme-Carosio, northwest Italy). Unfortunately, none of the foraminifera events are believed to be true stratigraphical tops (facies limited), and reworking in the section hampers use North Sea Oligocene-Lower Miocene dinoflagellates and acritarchs of nannofossil tops (Steininger et al., 1997). However, the dinoflagellate succession from the Lemme-Carosio section has been documented in detail by Powell (1986a), Brinkhuis et al. (1992) and Zevenboom (1995Zevenboom ( , 1996 and provides means for direct correlation between it and the North Sea Basin. The HO of Distatodinium biffii is below the Chattian-Aquitanian boundary in its type section and the HO of Chiropteridium spp. is above. This event succession can be recognized in many North Sea wells and the Chattian-Aquitanian boundary is positioned between the two. When positioned in this way, the Chattian-Aquitanian boundary is closely below the FDO of small forms of the diatom Aulacodiscus insignis quadrata, a widespread event in the North Sea.

Aquitanian-Burdigalian boundary
The principal criteria for the Aquitanian-Burdigalian boundary are undecided and the correlation is based on the dinoflagellate zonation scheme of de Verteuil & Norris (1996a), established   Table 1, underlined in Fig. 4) are shaded. 1 in the Series column refers to Eocene; 2 in the Stage column refers to Bartonian; R is assumed reworked; x is occurrence outside count. Sources: foraminifera, King (1989); dinoflagellates, Costa & Manum (1988), modified by Köthe (1990); de Verteuil & Norris (1996a); Mudge & Bujak (1996). for US East Coast sections. This dinoflagellate succession is calibrated directly with the international chronostratigraphy by calcareous nannofossils and foraminifera. De Verteuil & Norris (1996a) used the Miocene time-scale of Berggren et al. (1995) and showed that the Aquitanian-Burdigalian boundary is close to the HO of the dinoflagellate Caligodinium amiculum. Therefore, the top of the Aquitanian is placed at the FDO of C. amiculum, following de Verteuil & Norris (1996a).

RESULTS
Key species from Alma-1X can be used to correlate the well succession ( Fig. 2) with the zonation schemes of Costa & Manum (1988), Köthe (1990), Powell (1992), de Verteuil & Norris (1996a) and Mudge & Bujak (1996). None of the existing zonation schemes allows for a detailed subdivision of the study interval. However, the event succession in Alma-1X indicates that a number of conspicuous dinoflagellate and acritarch events may be of potential use in a detailed subdivision of the Oligocene-Lower Miocene. The event succession comprises primary key FDOs (Table 1, underlined in Fig. 4, shaded in Fig. 2) and secondary, supporting FDOs (shown without underlining in Fig. 4). The key events have been selected based on ease of identification and persistency in occurrence. The secondary events seem to occur less persistently than the key events in the well, but potentially may still be used as supportive correlatives.
At the Eocene-Oligocene boundary, in the basal part of the studied section, a hiatus is indicated by the dinoflagellate data. The two Priabonian marker events, FDO Heteraulacacysta Note: these taxa are shaded in Figure 2 and underlined in Figure 4.  . Key biostratigraphic events (underlined) and secondary, supporting events (without underlining) from the Alma-1X well (see text for further explanation). The correlation with North Sea microfossil zones is based on unpublished well data from Robertson Group plc. and King (1989).
porosa and FDO Rhombodinium porosum occur in the same cuttings sample as the FDO of Cerebrocysta bartonense, which marks the top of the Bartonian (upper Middle Eocene) according to Powell (1992) and Mudge & Bujak (1996). Hence, that sample is considered of Bartonian age, and the Priabonian is either lacking or incomplete in the well. On the seismic section and on the gamma log the hiatus is represented by the Top Horda Marker. Based on dinoflagellate biostratigraphy, the four seismic and log markers present above the Top Horda Marker can be dated as follows: Lark 4 Marker: intra-Rupelian; UOU Marker: uppermost Oligocene to lowermost Miocene; UL and Lark 18 markers: intra-Burdigalian (Fig. 4).

DISCUSSION
The present study demonstrates the potential for establishing a detailed subdivision of the Lower Miocene to Oligocene in the Central North Sea, based on FDOs of dinoflagellates and acritarchs. It is suggested that 26 key events may be of use in such a biostratigraphical subdivision (Table 1 and Figs 2, 4). The stratigraphical detail represented by the subdivision is comparable with that of the directly underlying Paleocene-Eocene section (Mudge & Bujak, 1996) and is much more detailed than established surrounding onshore zonation schemes. However, although the Lower Miocene-Oligocene of the Alma-1X well is considered fairly complete, hiati may be present in connection with the five seismic and log markers present in the interval. More studies of North Sea wells are needed to clarify this point and to establish a formal zonal subdivision of the interval.  Evitt, 1985), oval to sub-circular in outline. The number of wall layers could not be determined; the cyst probably has an autophragm. The cyst wall is fibrous and thick (3-7 µm), its surface is densely pitted. Numerous (>60), discrete, solid, fibrous, flexuous to anastomosing processes rise from the surface. The processes appear non-tabulate; however, they are often connected proximally by low ridges to form process rows or groups that may or may not reflect paratabulation. The processes may also be arranged in composite process groups (Figs 5c, d). Most processes have a broad base, but taper at half-length to become slender. Process shape varies on a specimen: ranging from simple and acuminate (Fig. 5a) over bifurcate to trifurcate (Fig. 5b) to licrate (Figs 5d, e, f). Some processes branch at half-length or above to form two or more sets of tips (e.g. Pl. 12, fig. 12 at c. 2 o'clock, fig. 13 at 9 and 10 o'clock; Figs 5e, g, h). The epicystal processes are usually slightly longer than those of the hypocystal area; the longest being those of the apex, the shortest being those in postcingular position. Process length variation may be from 5-15 µm on a specimen. The overall process length is fairly constant in a population. The archaeopyle is precingular, presumably P 3´´, judged from its shape (Pl. 12, figs 11-13). The paratabulation is difficult to determine and may be indicated only by the archaeopyle and, on some specimens, possibly by rows or groups of processes.

Stratigraphical range.
Middle to upper Rupelian in Alma-1X.  (Davey & Williams, 1966) Sarjeant, 1981 (especially specimens of that species figured by Jan du Chêne & Adediran, 1985, pl. 14, figs 3, 4, 9), but differs by having more and shorter, flexuous processes. Amphorosphaeridium? almae differs from A. multibrevum Davey, 1969, which also has numerous processes (although not as many as the new species), in having solid, flexuous processes that terminate bifid to licrate, sometimes branching, instead of having hollow processes that are spinous distally. Amphorosphaeridium? almae resembles species of Cordosphaeridium and Fibrocysta in its wall structure and in having fibrous processes, but differs from species of Cordosphaeridium in having numerous, non-tabulate processes instead of the much fewer and tabulate processes characteristic of the latter genus. It differs from species of Fibrocysta in lacking polar protrusions, and from species of Exochosphaeridium in having complex process tips instead of the simple process tips typical of the latter genus. Specimens of Amphorosphaeridium typically have hollow processes and possess dictictive polar processes (Davey, 1969). As these characters are not typical for the new species, it is placed in Amphorosphaeridium with some hesitation.
Genus Dinopterygium Bujak, 1984 emend. Head, 1994 Dinopterygium cladoides sensu Morgenroth, 1966 (Pl. 11, figs 11-12)  Diagnosis. Gonyaulacoid autocyst with an extremely thick autophragm consisting of a thin pedium and a very thick luxuria composed of a dense mat of slightly anastomosing fibrils that radiate outwards from the pedium. The archaeopyle is precingular, type P 3´´. The cyst lacks any indication of paratabulation other than the archaeopyle. Description. Small to intermediate-sized autocyst, sub-circular to oval in outline. The cyst wall is extremely thick and composed of two wall layers; a thin, smooth pedium (less than 0.5 µm thick) and a very thick luxuria (thickness: 8-12 µm). The luxuria is composed of closely spaced fibrils radiating out from the pedium to form a dense mat on the cyst. Each fibril seems to be slightly flexuous to anastomosing from its base to its tip (Pl. 6, fig. 20). The luxuria is finely granulate to pitted in surface view (Pl. 6, fig. 19). The archaeopyle is precingular, type P 3´´, judged from its shape. Apart from the archaeopyle, the cyst lacks indications of paratabulation. In a population, cyst size and wall thickness varies only very little, whereas the overall shape may vary from being circular to oval. When oval, the long axis may be either polar or equatorial, resulting in specimens being either higher than broad (Pl. 6, fig. 21), or broader than high (Pl. 6, fig. 16). The former shape is the most common by far.

Dimensions. Fifteen specimens were measured (µm).
Stratigraphical range. Chattian to lower Aquitanian in the Alma-1X well. Bujak, 1984emend. Head, 1994 and from most other small to intermediate-sized dinoflagellate cysts with fibrous cyst wall ornament and a type P archaeopyle in its much thicker luxuria. Some specimens of Tectatodinium pellitum Wall, 1967 emend. Head, 1994 may possess a thick cyst wall, having a superficial resemblance with the new species, but they differ in having an interconnecting, irregular, 'spongy' and lanate luxuria instead of the radiating, fibrous luxuria of F. pachyderma (see Head, 1994, for discussion of luxuria morphology in the genera Tectatodinium and Filisphaera).
Stratigraphical range. Uppermost Bartonian to lower Rupelian in Alma-1X. In Alma-1X, the FDO of the taxon is recorded at the same stratigraphic level as that observed by Lund (2002), Manum et al. (1989) and Williams & Manum (1999), i.e. in the lower Rupelian, below the FDO of Enneadocysta pectiniformis and above the FDO of Areosphaeridium diktyoplokum. Its occurrence pattern thus confirms the observation that the species constitutes an important intra-Rupelian stratigraphic marker (Lund, 2002).
Remarks. The prime reason for Lund (2002) establishing the new genus Pseudospiniferites was the 'tendency towards suture break-up around the primary single plate precingular archeopyle'. However, Lund also stated that these accessory  fig. 11). The line drawing shows the sexiform gonyaulacoid tabulation pattern. Interpreted plates in Kofoid notation, Taylor archaeopyle sutures occur only sporadically in a cyst population (Lund, 2002, p. 87). Three of his illustrated specimens are, indeed, ruptured cysts (Lund, 2002, pl. 1, figs 1, 2, 4, 7), one is a fragment (Lund, 2002, pl. 1, fig. 3) and two (Lund, 2002, pl. 1, figs 5, 6) are intact. The taxon is well known from North Sea boreholes, occuring in various states of preservation. The population encountered in Alma-1X occurs at the same stratigraphical level as the specimens described by Lund and is dominated by well-preserved specimens that do not show signs of accessory archaeopyle sutures. The specimens figured by Manum et al. (1989) from ODP Site 643 and by Williams & Manum (1999) from ODP Site 908 -both Norwegian Sea -are intact and also lack accessory archaeopyle sutures. Therefore, the character stressed as a criterion for establishing a new genus is believed to be a preservation artefact. The generic diagnosis of the genus Pseudospiniferites also includes the presence of granulate to (micro) reticulate surface ornament (Lund, 2002, p. 87). However, the emended diagnosis of the genus Spiniferites allows for such surface ornament (Sarjeant, 1970, p. 75) and several recognized species have an ornamented surface.
The specimens observed from Alma-1X, those figured by Manum et al. (1989), Williams & Manum (1999) and illustrated by Lund (2002;figs 5, 6) from the North Sea all have at least some flexuous processes. Although most processes on the holotype are straight, some are flexuous (Lund, 2002, pl. 1, figs 1, 2). As all the specimens mentioned above are undoubtedly conspecific, the description of the taxon is widened to allow for specimens with flexuous processes.

Spiniferites pseudofurcatus
Diagnosis. Skolochorate, spiniferate, gonyaulacoid cyst of intermediate to large size. The processes terminate bifurcate to tetrafurcate or ramify. Low septal crests connect the processes proximally. The periphragm is smooth, the endophragm is coarsely verrucate to vermiculate.
Description. Large gonyaulacoid autocyst, composed of a subcircular to oval central body and a ring-shaped membranous ectophragm. The central body is thin-walled and smooth, lacking indication of paratabulation apart from the angular precingular archaeopyle. The ectophragm is separated from the central body by nine, usually straight to slightly curved processes which are oval in cross-section at their proximal point of attachment, but elsewhere are flat with a thickened central ridge.
The processes appear almost evenly distributed on the dorsal surface of the central body close to its ambitus, giving the cyst a wheel-like appearance. It is not possible to determine the exact location of the processes, as the paratabulation of the species cannot be determined, except for the archaeopyle. The processes are probably attached on precingular, cingular and postcingular paraplates, as well as apically and antapically. For ease of recognition, the nine processes have been numbered 1-9 in a counterclockwise manner on the cyst positioned with its dorsal side up, with 1 being the process located directly above the archaeopyle (Fig. 7). When numbered in this way, processes 5 and 6 can be found in antapical position, 4 and 7 occupying postcingular positions, 8 and 3 cingular positions and 2 and 9 in precingular positions (Fig. 7). Distally, the processes are connected by a thin-walled ectophragm that curls over to the ventral side of the cyst. Ventrally, the ectophragm has a very large sub-circular opening that extends almost to the perimeter. The ectophragm usually has two large perforations in apical and antapical position, located above processes 1, 2 and 9 and 4-7 (Pl. 5, figs 1-3; Fig. 7). These two perforations are located on the dorsal side of the cyst and create a ribbon-like connection between the adjacent processes (between processes 1, 9 and 4, 5 in Fig. 7a; between 1, 2, 9 and 5, 6, 7 in Fig. 7b). Additional perforations may occur between the two large perforations (Fig. 7b). The archaeopyle is precingular, type P 3´´, from its shape and position. The paracingulum is indicated on some specimens by low ridges on the periphragm (Fig. 7a above process 8; Pl. 5, fig. 1). Apart from the archaeopyle and rare indications of paracingulum, the paratabulation is not indicated on the cyst.
Remarks. Manum et al. (1989, pl. 3, fig. 9) figured a specimen of a dinoflagellate cyst similar to Thalassiphora rota from ODP Site 643 in the Norwegian Sea as ?Invertocysta sp. 1. This taxon has nine processes connecting the cyst body with the ectophragm. The two taxa are considered conspecific. The taxon recorded by

Manum and collaborators is restricted to the Early Miocene
Evitosphaerula paratabulata Zone of Manum et al. (1989) and thus occurs in the same interval as T. rota. Zevenboom & Santarelli in Zevenboom (1995) figured two specimens of the manuscript species 'Dalella siciliense' from strata of Serravallian-Tortonian (Mid-Late Miocene) age in Sicily, South Italy. The taxon is similar to T. rota and the two taxa are probably conspecific. However, the stratigraphical occurrence interval for the Sicilian specimens is clearly above that of the specimens reported from the North Sea and the Greenland Sea. Powell (1986b, pl. 5, fig. 4;1986c, pl. 1, fig. 4) figured specimens similar to T. rota as Thalassiphora? sp. A from the Aquitanian and Tortonian stages of the Piedmont Basin, north Italy. The specimens are considered conspecific with T. rota. Head et al. (1989a, pl. 4, figs 1, 2, 4;pl. 6, figs 1-3, 6, 7) figured specimens similar to the new species (as Gen. et sp. indet. 1) from the Upper Miocene of ODP Leg 105, Labrador Sea. However, these differ in having fewer connecting processes than the new species and in being much smaller; only two-thirds of the size of T. rota. Specimens of Thalassiphora cf. delicata, an undescribed taxon from the Paleocene of the North Sea illustrated by Powell (1992, pl. 4.4, fig. 5), are similar to T. rota in being wheel-shaped and in having perforations in the periphragm. The new taxon differs from T. cf. delicata in having larger perforations and thinner processes.