Dinoflagellate cysts from the Oxfordian (Upper Jurassic) of Skye, Scotland and Southern Dorset, England

Ten sediment samples from Upper Oxfordian strata of Skye and six from the uppermost Lower Oxfordian of southern Dorset were processed for palynomorphs. Of the dinoflagellate cysts recognized, the genus Furzidinium and species F. sentum, Cleistosphaeridium sarmentum and Prolixosphaeridium floccum are described for the first time. The new combination Polygonium bavarica is proposed and Cleistosphaeridium ehrenbergii is retained.


INTRODUCTION
The modem era of Oxfordian dinoflagellate cyst research began in 1938 with the landmark paper by Deflandre on the microplankton of Villers-sur-Mer. northern France. Since then over sixty papers have been published (see Riley & Sarjeant 1972, Courtinat 1989 for references) with descriptions of Oxfordian dinoflagellate cysts and their stratigraphic ranges. This database is such that the present paper need only describe new fornis and comment on features which have not been reported previously. The other dinoflagellate cysts found in the study are listed in the distribution charts (Figs 1 1 -12, Table 1)

with selected examples being illustrated in Plates
For each of the species described the total body length, including the horn if present, but ignoring processes or any large ornament is given along with the breadth of the body. Cysts with an apical archeopyle have two length measurements: the specimens without an operculum (operc.) were measured separately from those with one. All size measurements are in micrometers (p) and, where possible, were rounded to the nearest integer. If more than 20 specimens of a species are recorded, the mean of the measurements is given in brackets between the range of values. This is only a guide and has no statistical significance due to the small sample size.
The rock samples used for this study are from Skye (Scotland) and southern Dorset (England). The section on Skye (Figs 1,2) was collected by the British Geological Survey and the ten samples (corresponding to slides GMUS E3 107 to GMUS E3 1 16 and specimen numbers MPK 9046 to MPK 909.5) are from the Amoeboceras rosenkrantzi zone of Late Oxfordian age. Wright (1973, fig. 3) mapped the locality  in his review of the Upper Oxfordian of Staffin, Isle of Skye. The six southern Dorset samples (corresponding to slides GMUS E2974 to GMUS E2979) were collected by Dr. J. K. Wright of the University of London from the Cardioceras cordatum zone, the upper ammonite zone of the Lower Oxfordian substage (Fig 3). A description of these sediments and a listing of the ammonites found at the locality (NG 700-818) was published by Wright (1986). The sediment was processed at the University of Saskatchewan using the standard fossil palynological preparation method. The organic residue was passed through 1 5 0 p and 1 2 p sieves with the 12-1 S O p fraction mounted. The resulting slides from each sample are marked after the sample number (e.g., "GMUS E3100/1"). Three hundred marine palynomorphs were recorded from each sample (Figs 1 1-12). The slides made from Dorset sediment are stored in the collections of the palynology laboratory, Department of Geological Sciences, University of Saskatchewan, Saskatoon, Saskatchewan. Those from the Skye sediment are stored at the British Geological Survey, Keyworth, Nottinghamshire.
Remarks. This species is very similar to Prolixosphaeridiurn capitaturn (Cookson and Eisenack 1960) Singh 1971 and difficulty was encountered in their separation. The diagnosis of P . capitaturn (Pl. 1,fig. 11) A does not include a reference to the possible alignment of the processes into rows. Singh (1971, p. 342-3) mentioned the lack of any process alignment on his specimen even though it is a prerequisite of the genus (Davey 1969, p. 160). The only recorded difference between P . capitatum and B . creberbarbata appears to be that the former is longer by 2 0 p . The latter was first described by Erkmen and Sarjeant (1980, p. 52-4) Courtinat: 166,pl. 12,fig. 5;pl. 13,fig. 12. Remarks. Courtinat (1989, p. 166) re-examined the plesiotype designated by Deflandre (1938) and gave a new diagnosis. He also decided to synonymise the species into anew species as he contended that the holotype was not designated, and a diagnosis was not given by Deflandre. While it is true that the name was published without strict adherence to the International Code of Botanic Nomenclature (I.C.B.N.) then in effect it is thought unnecessary to discard the name. Instead the diagnosis of C. ehrenbergii should be emended in the near future and a type specimen designated to conform with the I.C.B.N. presently in effect. The emended diagnosis of Courtinat (1989) is therefore rejected and the species Cleistosphaeridium deflandrei is considered to be a junior synonym of C. ehrenbergii. Derivation of name. Latin Sarmentum, twig or light branch, referring to the distal morphology of the processes. Diagnosis. Skolochorate cyst with a sub spherical body; wall smooth to slightly granular. Processes isolated, non tabular, 28-40 in number, cover the body and are long (half to third of body width), hollow, generally straight, gently tapering distally and have slightly granular walls. Proximally the processes flare at the contact with the body but do not communicate with it. Distally the processes flare briefly at their tips, appear closed and have up to seven distal spines radiating from the process tip. Archeopyle apical with or without an attached opercuium. Occurrence. In the present study, the new species was only found in the Skye sequence which is of Late Oxfordian Age (Amoeboceras rosenkrantzi Zone). The range could extend into the Early Kimmeridgian (Pictonia bavlei Zone) if the specimens found by Gitmez (1970) are assigned to the species. Remarks. These cysts have a morphology close to that of C.? tribuliferum (Sarjeant 1962) Davey, Downie, Sarjeant & Williams 1966. Gitmez (1970 included similar forms with a limited number of processes in the species but this is outside the range of the diagnosis as outlined by Sarjeant (1962, p. 487). A reexamination of these specimens would probably show that they can now be placed in C. sarmentum. No evidence of intergradation between these two species has been observed in the present study. The species is differentiated from C.? tribuliferum by its smaller number of processes, more complex distal spine morphology and the relatively shorter process length as compared to the body width. No other species of Cleistosphaeridium has a similar combination of features. Hvstrichosphaeridium petilum Gitmez 1970 has fewer processes which are intra tabular.
Furzidinium gen. nov. Derivation of name. Named after Furzy Cliff from which the type species was first reported.

Explanation of Plate 1
Figs. 1-2. Cleistosphaeridium sarmentum sp. nov. Holotype. Slide GMUS E3113/2, England Finder coordinate H37/0, MPK 9047. Lateral view, phase contrast, magnification circa x 1 150. Fig. 3; Pilosidinium echinatum (Gitmez and Sarjeant 1972) Courtinat 1989. GMUS E2974/1, England Finder coordinate C35/0. Dorso-ventral view, phase contrast, magnification circa x650. Fig. 4. Furzidinium sentum sp. nov. Paratype 1. Slide GMUS E2976/3, England Finder coordinate C32/0. Lateral view, phase contrast, the epicyst with a well developed mamelon has slightly rotated into the hypocyst. Magnification circa x1200.   (Sarjeant 1962); emend. Davey 1979. Slide GMUS E3 108/2, England Finder coordinate H38/4, MPK 9063. Dorso-ventral view, phase contrast, magnification circa x650. Fig. 7. Cfeistosphaeridium ehrenhergii (Deflandre 1947) Davey, Downie, Sarjeant and Williams 1969. Slide GMUS E2977/1, England Finder coordinate B46/ 3. Lateral view, phase contrast, magnification circa x650. Fig. 8. Cleistosphaeridium? trihuliferum (Sarjeant 1962) Davey, Downie, Sarjeant and Williams 1969. Slide GMUS E2976/1, England Finder coordinate E3 1/ 1. Lateral view, phase contrast, magnification circa x650. Fig. 9. Systematophora valensii (Sarjeant 1960) Downie and Sarjeant 1964. Slide GMUS E3109/1, England Finder coordinate R38/4, MPK 9066. Lateral view, phase contrast, magnification circa x650. Fig. 10. Barhatacysta cf. hrevispinosa (Courtinat 1980 (17) Remarks. In the assemblages studied, the species was found both with and without the epicyst though the former state is much more common. The larger processes are gonal, and there is occasionally a bifurcation at their base into two long processes. Intergonal processes are best, though not exclusively, developed at the contact of the apical and precingular plates on the epicyst and at the contact of the postcingular and antapical plate. Processes are developed on the hypocystal side of the cingulum but were not observed on the epicystal side. The wide cingulum is marked by low walls of equal height which are larger than those marking the sutures. The operculum is probably adherent in nature. Where the operculum is not retained, the edge of the archeopyle is smooth around the cyst without the development of a sulcal tab projecting from the edge of the opening (see P1. 3,Figs 4,9). This feature makes the interpretation of the sulcal area unique but the lack of a sulcal tab combined with the well developed cingulum precludes other more classical interpretations. The apical mamelon is always present; its tip does not accept stain and under phase illumination appears darker than the rest of the apex.  Sarjeant 196 l a Remarks. The illustrated specimen from the southern Dorset sequence does not have a well developed archeopyle but does have an extended apical structure. This is 'rod like' and similar to those developed by Pareodinia ceratophora Deflandre 1947. A second specimen was found in the southem Dorset sequence which is compressed vertically though a similar apical horn extension is preserved.  (1988, p. 348-351) but the generic diagnosis outlined by Fensome (1979. p. 50-51) requires constituent species to be cavate and have a well developed paratabulation. P . bavarica does not have a resolvable paratabulation and is therefore transferred to Polygonifera. The inclusion of Pocock's (1972) record in the synonymy listing is based on his photograph. Jansonius (1986, p. 202) reported that the specimen is now lost. The type material of this species was recorded from the Middle Oxfordian of Bavaria. The range of the species is now increased to Explanation of Plate 2  (Sarjeant 1963 (Deflandre 1938    Lateral view, phase contrast, magnification circa x500. Fig. 9. Furzidinium senturn sp. nov. Paratype 2. Slide GMUS E297812, England Finder coordinate F43/2. Lateral view, phase contrast. The hypocyst showing the smooth cingular margin without a sulcal prominance. Magnification circa x750. Fig. 10. Leptodinium ambigiturn (Deflandre 1939) Sarjeant 1969. SlideGMUS E3 1 13/1 ,England FindercoordinateE42/1 ,MPK 9080. Lateral view, phasecontrast, magnification circa x600. Fig. 1  include the top of the Oxfordian of Skye. Pocock did not record in detail the strata from which he found his specimen. Prolixosphaeridium Davey, Downie, Sarjeant and Williams 1966; emend. Davey I969 Remarks. The type species was originally erected by Deflandre (1 937) as a sub species of Hystrichosphaeridiurn xanthiopyxides (0. Wetzel 1933) Deflandre 1937, with a very brief diagnosis ... "An individual, ellipsoidal, punctate, covered with many short spines" (translation from French). No dimensions were given and there is not a photograph in the paper; however, a scale line diagram was presented and it can be used to estimate the size of the cyst. Central body length, with archeopyle circa 4 0 p , breadth circa 2 0 p , process length 4 -7 p , number of processes 80-100. This broadly agrees with the dimehsions given by Sarjeant (1978, p. 23). Davey & Verdier (1974, p. 636-7) designated the original type species, Prolixosphaeridium deirense Davey, Downie, Sarjeant & Williams 1969 as a junior synonym of P.pawispinurn but did not redescribe or illustrate the type species. A restudy of the original type specimen of P . parvispinum and the designation of paratypes is urgently required, if it is to be of use as the type species of Prolixosphaeridium. The difference between Prolixosphaeridium and Sentusidinium Sarjeant & Stover 1978 is that the former has non tabulate processes arranged in lines extending around the ellipsoidal body of the cyst. This has lead to problems with species which have many small processes, not arranged in a discemable pattern, and only somewhat elongate central bodies; for example Prolixosphaeridium capitatum (Cookson &  Diagnosis. Elongate ellipsoidal proximochorate cyst, its surface covered with coarse granules. Processes sparse (circa. 40-50 on cysts without attached archeopyles) flexible and fairly long (up to a third of the cyst's breadth); they can be flared at their bases, tapering distally, straight, synodal or bent at right angles (especially the processes near the antapex). Process tips are closed, typically acuminate, rarely blunt. Tabulation indicated by the six plates of the archeopyle and on well preserved specimens a sulcus is discemable. A faint cingulum may also be indicated by two lines of enlarged granules between which is an area devoid of processes.      (Sarjeant 1962) fig. 13) Remarks. The differentiation of this species from Adnatosphaeridium caulleryi (Deflandre 1938) is not simple, despite the emendation and remarks in Stancliffe & Sarjeant (1 990). Systematophora orbifera has well organized processes forming distal rings while, at its most developed, A . caulleryi has major process clusters only on the precingular, post-cingular and antapical plates. It is apparent that the two species are closely related morphologically with A. caulleryi possibly being the ancestral form. secondary processes, and the development of a discernable cingulum and sulcus is unique to this species.

Scriniodinium Klement 1960
Scriniodinium crystallinum (Deflandre 1938) Klement 1960 (PI. 2, fig. 7; text- fig. 9) Remarks. The apical horn is not always present and can be just a slight bulge in the periphragm. The internal body may, or may not, have a slight development of an apical horn and no opisthopyles were observed on the inner body.
The specimen drawn and photographed shows a well developed enlarged peripyle which includes a cingular plate. This archeopyle is also unusual as it extends apically to include another plate. This plate could be part of the apical series. The horn has also separated from the cyst. A similar specimen was illustrated by Klement (1960, plate 1, figure 1). The endopyle is smaller than the peripyle and not so well delineated. Fig. 9. Scriniodinium crystallinum (Deflandre 1938) Klement 1960. Slide GMUS E3 1 12/1, England Finder coordinate G30/0, MPK 9053. Dorsal (lower) surface. The secondary opening above the 3" plate is well shown as is the loss of a small secondary plate from the apical horn. Magnification x2100. (Sarjeant 1960) Sarjeant 1961 (Pl. 1, fig. 9; P1.4, fig. 9) Remarks This species appears closelyr elated to Systematophora areolata Klement 1960 and Systematophora penicillata (Ehrenberg 1843 ex 1854) Sarjeant 1980 and probably has an intergradation of morphological features. The raised wall linking the process bases of a cluster and the two or more stages of distal branching of eachprocess distinguisyh the species from S. areolata. S. penicillata has the presence of paired processes marking the cingulum which is not the case with S. valensii. Tubotuberella Vozzhennikova 1967;emend. Sarjeant 1982 Tubotyuberella apatela (Cookson & Eisenack 1960); emend. Sarjeant 1984 (PI. 3, figs. 11-12;text-fig. 10

2'
Fig. 10 A-F. Variation in the shape of the antapical extension of the periphragm of Tuhotuberella apatela (Cookson & Eisenack 1960 Remarks. Sarjeant (1982. p.42), following his emended diagnosis of this species, notes in a discussion section the difference between Tuhotuberella apatela and Tuhotuherella dangeardii (Sarjeant 1968) Stover and Evitt 1978. The former has a more elongate outline and a hypopericoel (hypo.) whose length is equal to or slightly greater than its breadth. The width of a hypopericoel is difficult to measure as it varies along its length. The length of a hypopericoel was also not classified in the paper and this causes problems if the separation of the two walls of the cyst extends along way up theendoblast towards the cingulum. Text- figure 10 shows the suggested measurements used in the present study: the distance from the antapical tip of the inner body (length A 2) versus the width (Ant 2) of the hypopericoel at the antapical tip of the inner body, for the specific differentiation. This removes the ambiguity of the classification and also minimizes the variability of laboratory processing which can alter the amount of wall to wall contact. This effect was discussed and illustrated by Schrank (1988) in a paper on the classification of Cretaceous Peridinioid species but it is probable that all two and three walled cysts behave the same way.

DISCUSSION OF THE ASSEMBLAGES STUDIED
The two sequences described contain diverse assemblages which are recorded in Text-figures 1 1 -12. The study recorded 46 species of dinoflagellate cysts which have been assigned to 39 genera. The Skye sequence has 32 genera and 35 species of dinoflagellate cysts and the southern Dorset sequence has 32 genera and 30 species of dinoflagellate cysts. There have been many published and unpublished papers on the biostratigraphy of British sequences of Oxfordian Age, and detailed analysis of all these reports is beyond the scope of the present study. However, Williams & Bujak (1985) summarized the important dinocyst occurrences, listing 2 1 species of dinoflage1:ate cysts which have biostratigraphic significance in the Late Oxfxdian. Of these, this study found in the Skye sequence Pareodinia cei-atophora, Ellipsoidictyurn cincturn, Gonyaulacysta jurassica, Tubotuhei-ella apatela, Sirrniodiniurn grossii, and Scriniodiniurn crystallinurn. For the Early Oxfordian, Williams & Bujak (1985) listed 32 dinoflagellate cyst species of biostratigraphic significance of which Riqaudella aernula, Pareodinia ceratophor-a, Nannoceratopsis pellucida, Sirrniodiniopsis orhis, Gonyaulacysta jurassica, Neti-elytron stegasturn, Tuhotuherella apatela, Sirrniodiniurn gi-ossii, Stephanelytron spp., Atopodiniurn prostaturn and Scriniodiniurn crystallinurn were found in the southern Dorset sequence. A search of the other recorded species in this study found only one anomalous occurrence (as outlined by Williams & Bujak, 1985),Senoniasphaera jurassica, which has been previously reported to have its first Occurrence in the Early Kimmeridgian. In this study the sole recorded specimen was found in the Early Oxfordian; further examples have to be recorded before an extension in the range of the species can be contemplated.  The paleoenvironment of deposition can be inferred from the assemblages already described. Both sequences have a high diversity of dinoflagellate cysts which, according to Williams and Sarjeant (1967, p. 393) indicates an open marine environment with normal levels of salinity. The chorate cysts only form a small part (generally less than 7%) of the assemblages which is the norm in 2 other studies of Oxfordian dinoflagellate cysts (Sarjeant 1968, tab. I). In a study of Cenomanian marine palynomorphs, Davey (1970, p. 395) found that cavate dinoflagellate cysts dominate in cool water while in warm water they are rare. Though these cysts are peridinacean forms, it is thought that gonyaulacean cavate forms have such a similar overall morphology that comparable reationships to water temperature can be tentatively inferred. Davey also suggested that chorate cysts should be more abundant in warm seas due to their ability to sink in the less dense water. In the present study, cavate andchorate cysts are always a minor component of the assemblages. However, no paleoenvironmental conclusions can be made as most Upper Jurassic assemblages are dominated by proximate cysts. This is, no doubt, influenced by the evolution of cyst morphology which initially was dominated by proximate cysts in the Triassic and Lower Jurassic. Cavate and chorate cysts only started to become the dominant morphologies in later assemblages such as those from Cretaceous sediments. This evidence indicates that the two sequences are both from sediments deposited in fully marine environments which were similar, though not identical. The marine environment was not of excessively high energy, and had no extremes of salinity or temperature which could have inhibited the growth of marine palynomorphs or the preservation of their cysts upon deposition in the sediment. The large number of terrestrial pollen and spores found (though not subjects of detailed study here) indicates aproximity to land. The paleogeography of the area was described by Cariou et al. (1985) using invertebrate fossils and this agrees well with the interpretation of the two sequences studied.