Review of the dinoflagellate cyst Subtilisphaera? inaffecta (Drugg, 1978) Bujak & Davies, 1983 and S.? paeminosa (Drugg, 1978) Bujak & Davies, 1983

Research carried out on the Upper Jurassic dinoflagellate cyst assemblages of the Sub-Tethyan marine realm, show that populations of the dinoflagellate cysts Subtilisphaera? inaffecta and S.? paeminosa are predominant in shallow water marginal marine or brackish environments. The distribution of groups of dinoflagellate cysts, micrhystridid acritarchs and variations of terrestrial inputs represented by phytoclasts are presumed parameters of the salinity balance during such Late Jurassic depositional environments. In this context, the shagreenate to faintly granulate S.? inaffecta appears to be an opportunistic taxon with an ability to prosper in brackish environments. In contrast, the coarsely granulate to pustulate paeminosa form is seemingly less eurytopic and flourishes with success in shallow, marginal marine, environments. SEM studies reveals that the two morphotypes possess transapical archaeopyle sutures on what is usually considered the antapex. Following these observations the cysts are interpreted in a reverse sense. Consequently, the attribution to the genus Subtilisphaera becomes inappropriate. The two morphotypes, interpreted as variants of a single species, are attributed to the genus Corculodinium Batten & Lister, 1988 for which a new emendation is proposed. The specific epithet inaffecta is considered legal over paeminosa.


INTRODUCTION
Research carried out on the Kimmeridgian and Tithonian dinoflagellate cyst assemblages in the Sub-Tethyan (Sub-Mediterranean) marine realm, record populations of Subtilisphaera specimens that are commonly predominant. The two Late Jurassic species are S.? inafecturn and S.? paerninosa, both erected by Drugg (1978). The Subtilisphaera populations described herein come from the Loches section (Aude) and the Solen Well (Lot); Fig. 1. These two sections yielded diverse dinoflagellate cyst and sporomorph populations plus large amounts of phytoclast debris or amorphous organic matter (AOM). Phytoclast particles are dominant in the Kimmeridgian samples and AOM is occasionally abundant in the Tithonian samples. Miospores are, in general, well preserved, presumably due to their more robust nature compared with the thinnerwalled dinoflagellate cysts. The latter show symptoms of physical degradation; they are colourless and residues are commonly composed of fragmented dinoflagellate cysts. Skolochorate cysts markedly are exceptionally well preserved, most specimens having damaged or twisted processes.
The studied material has been obtained applying standard preparation techniques. Quantitative palynological analysis of the overall kerogen composition was carried out on sieved unoxidized material and without application of ultrasonic methods. A series of traverses across the slide to reach a standard number of 220 phytoclast particles made the kerogen counts. In counting the 220 phytoclast particles all the observed palynomorphs registered [maximum 506, average count 266, minimum 1561. The data generated by the counting procedure are relative percentage particle abundances. Slides are housed in the University Claude-Bernard collections.
The objectives of the present study are to review the Kimmeridgian species formerly attributed to the genus Subtilisphaera, to discuss their palaeoecology and to present a new point of view on their systematic attribution.

PALAEOCOLOGICAL INTERPRETATIONS
In term of relative abundance the upper Jurassic dinoflagellate cyst assemblages studied (Appendix) present obvious differences between both northern and southern European assemblages. Diirr (1988) has established that Kimmeridgian dinoflagellate cyst microfloras of southern Germany (central Tethyan realm) are dissimilar in overall species content to those from more northerly (Boreal) areas but not in relative abundance. For example, proximate (Apteodinium, Cribroperidiniuni and Kallosphaeridium groups), proximochorate (Burbatacysta and Epiplosphuera groups) and skolochorate cysts (Systenzatophora group) are abundant and diverse in the Lower Kimmeridgian of Germany as well as in the British Isles. All these forms are cosmopolitan species.
In common with European assemblages (Boreal and Tethyan realms), species of Subtilisphaera in the studied sections proved to be dominant constituents in some instances (up to 65% of palynomorphs; 16% of kerogen). Furthermore, the Micrliystridium group were occasionally present in obviously significant proportions (up to 40%) together with the Prusimphyceae group (up to 25% of palynomorphs). The abundance of cavate cysts, especially forms of the genus, Subtilisphaera have been reported in the upper part of the Eudoxus ammonite Zone of Quercy by Ioannides et al. (1988). Diirr (1988) noticed the abundant population of S.? inageeta and S.? paeniinosa in the Late Kimmeridgian (2 Zone) of Germany and Poulsen (1996) also mentions a rich flora in the Late Kimmeridgian of Denmark and Poland. Poulsen (1996) emphasized the particular environment that favoured the abundance of Subtilisphaera forms. They appear to indicate shallow water, and probably restricted, low energy depositional environments.
paenzinosa. The Subtilisphaera population presents maxima of abundance associated to minima of the dinoflagellate cyst:spor-omorph ratio (Fig. 3) and low dinoflagellate cyst diversity (Fig.  3). The dinocystic : sporomorph ratio generally declines sharply onshore and is especially low in deltaic areas or above the tidal limit (Tyson, 1995). These patterns reflect partial dilution by the high sporomorph and sediment influx. In the S.? inafectu maximum of the Loches section, grains of pollen identified as Spher@ollenites belonging to the Cheirolepidacean-Taxodiacean conifers are over-dominant. Mesozoic inhabitants of swamp environments comparable to mangroves produce these pollen. The Szkhri/is(thneva population maxima of the Loches section probably reflects the most nearshore depositional environments for the samples we have analysed. The acanthomorph acritarch Mici-h,vstridium is well-represented and is occasionally abundant (Fig. 3). Many studies have regularly established that a relative abundance of small micrhystridid acritarchs is most representative of shallow water marginal marine conditions (Erkmen & Sarjeant, 1980;Schrank, 1984;Prauss, 1989). In the Loches section, samples M03, M07-M 0 8 and MI3 show an abundance peak of short spined Micrhystridiunz. The M03 and M07 peaks are associated with a slight increase in the percentages of the elongated or needle-like phytoclasts (PM lT, PM4Tm and PM4Tn; Fig. 3). These palynomacerals are presumed to possess an elevated buoyancy and are deposited under low energy conditions (Courtinat & Rio, in preparation). The M13 peak is associated with high relative frequencies of phytoclasts (exclusively the PMlE particles), an increase in the relative abundance of the prasinophytes and a low dinocyst diversity. As for the precedent peaks, the M13 'event' appears related to brackish marginal facies. This contention is supported by the general trend of decreasing energy conditions, relatively large inputs of phytoclasts induced by freshwater arrivals, and decrease in the dinoflagellate cyst diversity.
In the Loches section the Prasinophyceae group shows an increasing richness from base to top. This group is known to be an important element of the modern phytoplankton in high latitude seas (Bird & Karl, 1992). A general association or affinity with cold waters for fossil prasinophytes is questionable. Most, but not all, of these occurrences are reported in what are interpreted as dysoxic-anoxic facies. The samples we have analysed from the Loches section show no discernible dysoxicanoxic event. The pyrolysis data and the absence of true AOM support this assumption. Many workers have regarded the fossil prasinophytes as indicators of brackish surface water conditions. Tyson (1 995) demonstrates that the brackish hypothesis is an apparent general agreement based on an uncritical acceptance of earlier interpretations or on the brackish and freshwater occurrence of some modern motile prasinophytes. His recommendation is that this hypothesis is not applied to all occurrences. The brackish hypothesis presumes major freshwater inputs in the marine surface layer. In this case, prasinophytes are associated with high sporomorph and phytoclast contents. These trends are observed in the Loches section for the phytoclasts (Fig. 3) and for the sporomorphs; although the opposite hypothesis of a hypersaline environment has been assumed (Bernier & Courtinat, 1979;Kunz, 1990). It is realistic to suggest that the prasinophytes are opportunistic forms occupying ecological niches deserted by phytoplankton coinmunities when ecological stress of variable nature occurs.
Review of the dinoflagellate cyst Subtilisphaera? inaffecta and SY paeminosa

The Solen well
The Tithonian samples examined from the Solen Well (Gigas Ammonite Zone; Quercy; Figs 1, 2) yielded low diversity assemblages of dinoflagellate cysts and sporomorphs plus large amounts of phytoclast particles (mainly abundant in pure limestones) and amorphous organic matter (AOM; chiefly abundant in marly limestones). The Subtilisphaera population represents 65% of the palynomorphs (Fig. 4). The black, equidimensional, polygonal woody particles are common in all lithologies. The relationship between the variations in the Subtilisphaera population and other data is different from that observed in the Loches section. The Subtilisphaera peaks are unrelated to palynomaceral particles, dinocyst diversity or any other palynofacies parameters. This is due to the particular context favouring deposition of AOM (Courtinat & Hantzpergue; in prep.).
Gradation between the two AOM types observed, sheath-like and diffuse-edged or well-delimited aggregates, and woody

Om-
particles indicates that the AOM aggregates are amorphous terrestrial humic matter. High frequencies of Circumpofles, low and fluctuating dinoflagellate cyst and spore diversities and quasi-absence of bisaccate pollen are in agreement with the suggestion of a nearshore environment, subject to salinity variations and tidal-wave influences where continental run-off alternates with shallow marine conditions. The role of bioturbation, the diurnal-nocturnal activity of cyanobacterian communities and their trapping effect can explain the weak oxidation and, ultimately, the preservation of the sedimentary organic matter. Rhythms marked by palynofacies devoid of aggregates in association or not with sheath-like AOM are observed. These unequal rhythms are thought to be related to the vigour of oxic degradation and salinity balance, depending on turbulence and desalinization events of the surface waters by wind-driven currents, tides and run-off and presence of bioelements favoured by salinity and water depth modifications. The variations in the Jurassic Subtilisphaera population are presumably signs of salinity modifications.
In conclusion, the whole of the data show that the late Jurassic Subtilisphaera-rich depositional environments were shallow water marginal marine or brackish. Variable distribution of groups of dinoflagellate cysts, micrhystridid acritarchs and prasinophytes are associated with variations of terrestrial inputs represented by phytociasts that are supposed parameters of the salinity balance prevailing during the Eudoxus Ammonite zone at Loches section and the Gigus Ammonite Zone of the Solen well. In this context, S.? inufecta appears to have a propensity to flourish in brackish waters (low salinities) and S.? paeminosa to favour increased marine waters (normal salinity). The two forms are assumed to be two variants of a single morph (ecoplienotypes). The shagreenate or weakly granulate morphotype is presumably eurytopic with a good adaptability for brackish environments. The coarsely granulate sculptured morphotype is seemingly stenotopic and an inhabitant of shallow marginal marine environments. The possibility that the laevigate/granulate ratio could be a palaeosalinity marker is important. The concept of ecophenotype rarely receives attention in palynological studies. Monteil (1991), in a very detailed work dealing with the plexus MuderongialPhoherocysta, proposed for each so-called smooth ecotype (Muderongiu species) four ecotypes with processes (Phoberocysta species). Smooth species (Muderongiu) are dominant in outer shelf sediments, whereas forms bearing processes (Phoberocysta) are more abundant in inner shelf sediments. In a similar way, Feist-Burkhardt & Pittet (1996) have considered that the increase or decrease of surface ornamentation of some Jurassic forms attributed to the MendicodiniumlCtenidodinium complex are the result of environmental factors. Drugg (1978), during a Jurassic sampling programme to gain stratigraphically controlled material for palynological research, describes two new cavate forms in the Kimmeridgian of England and Germany. The major particular attribute of these species remains the difficulty in recognizing an archeopyle.

SYSTEMATIC AND MORPHOLOGICAL BACKGROUNDS
In the opinion of the author, the two species are closely related. Both are primarily epicavate to bicavate cysts with a pentagonal to elliptical outline. The pericoel can be restricted at one extremity where the periphragm forms a bulge (an apical Review of the dinoflagellate cyst Subtilisphaera? inaffecta and S? paeminosa T horn from the Drugg's assumption conferring the aspect of an epicavate cyst) or at the two opposite extremities (bicavate cysts; with interpreted cone-shaped apical horn and roundish to subangular antapical horn). Endocysts are circular to elliptical, smooth-, granulate-or pustulae/verrucae-walled. The only perceptible paratabulation elements are two low transverse ridges delineating a paracingulum. Bujak & Davies (1983) transfer the two species to the genus Subtilisphaera. These new assignments are based on the interpretation of a transverse archeopyle (Bujak & Davies, 1983: 'Specimens illustrated by Drugg [1978] as S. inafleecta and S. pneminosa from the Kimmeridgian of England and Germany, ... appear to have the same transverse archeopyle type.'). The expression 'the same transverse archeopyle type' refers to a transverse archeopyle suture present between the apical and intercalary series developed on cysts with a bipesoid paratabulation. This interpretation is not accepted by Lentin & Williams (1985) as the illustrations of Drugg (1978;pl. 3, figs 5-12) show neither evidence of transverse archeopyle nor indication of a bipesioid paratabulation. They questionably retained the two species in Subtilisphaera. For most authors, Lentin and Williams are the authority on this subject. Evitt (1985) suggests that the genus Subtilisphaera is an archeopyle-less Alterbidinium Lentin & Williams, 1985 emend. Khowaja-Ateequzzaman er al., 1991.
All these opinions postulate that the epitheca is bell-shaped and that the hypotheca possesses one or two undeveloped antapical horns. These points of view converge towards an apparent consensus based on an uncritical acceptance of a standard morphology of cysts looking like the P-Cysts of Evitt (1985), i.e. cysts that would be referable to living Peridinium, Protoperidinium, or closely similar genera on the basis of their morphology. However, in this case, neither S.? inafecta nor S.
Review of the dinoflagellate cyst Subtilisphaera? inilfjecta and S? paenzinosa ?paeniinosa present a full morphology that looks like a peridinoid one.
Earlier suggestions of a different interpretation rely on studies of the Kimmeridgian and Portlandian strata of Jura where Subtilisphaera forms are sometimes dominant in the palynomorph assemblages. As a rule, Subtilisphaera specimens are commonly damaged on what is the antapex in the current interpretation. Following these observations the suggestion that the cysts could be interpreted upside down has been postulated. A presumed relation with the genus Saeptodinium Harris, 1974 has been suggested but without solid argument (Courtinat, 1989). Recent SEM observations render this conjecture (interpretation of the cysts in reverse sense) worthy of further consideration.
Although Lentin &Williams (1985) recorded S. inujbcta and S. pacminosa, with some doubts in the genus Subtilisphaera, their emendation of this generic taxon is questionable because of the nature of the archeopyle of the type species.
The original diagnosis of the genus Suhtilisphuera Jain & Millepied, 1973 (p. 27) is: 'Shell pentagonal-ovoid, test cavate to bicavate non-tabulate, asymmetrical; cingulum (girdle) well developed, dividing the shell into almost equal halves. Epitract broadly rounded with pointed to broadly obtuse apical horn; hypotract rounded having one prominent antapical horn and second undeveloped or only as a slight projection placed away from median axis (non-axial). Periphragm smooth to granulate, thin, delicate. Endophragm well developed, smooth, thin, delicate; capsule circular, filling periphragm completely, or leaving a small pericoel. Archeopyle mostly not seen, if present intercalary.' The emended diagnosis of Subtilisphaera by Lentin & Williams (1 976) reported the transverse archeopyle (?AIP) resulting from separation along a transapical suture. For the authors, the operculum, which may include paraplates 3', 1-3a and 3"'-5", remains attached along the posterior parasuture (operculum adnate). The type species Suhtilisphaera senegalensis Jain & Millepied 1976 shows no archeopyle. Stover & Evitt (1978) and Below (1981) have questioned the use of such a negative character and have stressed the uncertainties of the catch-all Suhrilipshuera genus. Stover & Evitt (1 978) assumed the distinctions of the Geiselodinium Krutzsch, 1962, Subtilispt'zaera, Saeptodinium and Teneridiniunz Krutzsch 1962 genera are disputable. The opinion of Evitt (1985) is to consider that Geiselodinium and Teneridiniunz genera are apparently freshwater cysts, whose thin and folded walls leave details of morphology quite unclear, while Subtilisphaera is a more distinctive marine cyst. The consideration of the mode of life in the distinction of undiscriminating peridinoid cysts is a possible way forward that merits attention.

SEM OBSERVATIONS -SYSTEMATIC IMPLICATIONS
Observed under transmitted light microscope or in SEM, 70% of the studied Subtilisphaera forms present a break (Pl. 1, figs 1,  3, 4 and 6). Under transmitted light microscope this fracture is not interpretable. By contrast, SEM analysis shows that the rupture is not a mechanical split but probably has a structural origin because of the presence of accessory sutures (PI. I , figs 6 and 8). The principal suture extends adapically from one side of the paracingulum to the other and traverses the apex (Pl. 1, figs 1  and 4). This tear is interpreted as a transapical archeopyle suture. This interpretation renders the half valve that supports the archeopyle suture equivalent to an epicyst. By opposition, the horn would be antapical in position. In addition to the archeopyle suture and the associated accessory sutures, the paracingulum and some paraplate sutures complete the paratabulation. Paraplate sutures are perceivable on the paracingulum (PI. 1 , figs 2 and 7), the apex (PI. 1, fig. 2) and the antapex (PI. 1, fig. 7). On some specimens an opisthopyle is present (Pl. 1,  fig. I). The laevigate to faintly granulate inafecta morph (PI. 1, fig. 4) grades into the more coarsely ornamented paerninmu morph (PI. 1, fig. 5). The separation of the two ornamentation types is not appreciated under SEM. The retention of the two distinct taxa seems irrelevant since the ornamentation is the major criterion of differentiation. The forms that exhibit a coarse ornamentation generally have a more polygonal epicyst and cone-shaped hypocyst (PI. 1, figs 2 and 7). By comparison, those with the lesser-ornamented periphragm usually have a circular epicyst and a pointed antapex (Pl. 1, fig 9).
The SEM observations, confirmed by those performed under transmitted light microscope, that indicate the orientation of the two forms S.? inuffbcto and S.? paeminosa has been misinterpreted. The more probable exact orientation, instituted on the common assumption that all the transverse archeopyles are epicystal, is to consider the horn as an antapical one. As a consequence, the two morphotypes have characters that are not
Corculodinium uniconicum Batten & Lister, 1988 inaffectum (Drugg, 1978)   those of all the suggested genera (principally Geiselodinium and Subtilisphaera and, to a lesser degree, Saeptodiniurn and Teneridinium). Considering the palaeocological adaptation and the diagnostic characters of the morphotypes, the most adequate genus is Corculodinium Batten & Lister, 1988. To accept a transverse archeopyle, forms with two layers, a well-expressed paracingulum, enlarged diversity of wall ornamentation and occasional presence of an opisthopyle, an emendation of the genus Corculodinium is proposed. Original description. Small, proximate, heart-shaped cyst; epicyst without apical horn or prominence, usually slightly indented at apex when undehisced, broader than hypocyst which is typically conical. Phragma thin, consisting of two closely adpressed layers, sometimes with a minute antapical pericoel, but periphragm may not be developed in some specimens; when present it is laevigate and endophragm is scabrate to minutely granulate. Paratabulation either undiscernible or partly indicated by archeopyle development and sometimes by crumpling of periphragm, interpreted as peridiniacean. Archeopyle, com-3h-k. bination apical/intercalary (tAtI), formed by partial dehiscence of simple operculum which is adnate ventrally, sometimes enlarged by secondary splitting at aperture margin along accessory sutures between some precingular paraplates. Emended description. Hypocavate to bicavate, unusually circumcavate, circular to heart-shaped cyst. Periphragm and endophragm are frequently closely appressed and when a pericoel is present, the cavity is feeble. Epicyst without apical horn but occasionally with one or two weakly developed bulges. Epicyst broader than the conical hypocyst which typically and constantly presents a low roundish to sub-angular antapical horn. Endocyst circular, elliptical or heart-shaped. Wall surface shagreenate, scabrate, granulate, pustulate or verrucate. Periphragm shagreenate, scabrate or faintly and finely granulate to pustulate. Paratabulation either undiscernible or partly indicated by archeopyle development, sometimes by crumpling of periphragm or two low transverse ridges delineating a level (planar) or weakly levogyre paracingulum. When observed under SEM, some specimens show paraplate sutures but the paratabulation cannot be fully defined. Archeopyle variable, transapical (tAtItP)a/? or combination apical/intercalary (tA-tI)a/(tAtI)a, formed by partial dehiscence of simple operculum which is adnate ventrally, sometimes enlarged by secondary splitting at aperture margin along accessory sutures between some precingular paraplates.

Dimensions.
Total length 23 to 35 pm. Stratigraphical range. See Table 2. Remarks. The original description is enlarged to include: (1) the varibility of the archeopyle; (2) the disconnection of the periphragm and endophragm while formerly Corculodinium is described for typically acavate forms; (3) the presence on some specimens of a paracingulum and an opisthopyle; and (4) the extended palette of the ornamentation types of the layer surfaces which evolved from laevigate to pustulate-walled. I Kimmeridgian I Tithonian 1181  Table 2. Remarks. Corculodinium inaffectum possesses an interpreted transapical archeopyle, no apical horn and a central antapical horn. These features are those neither of the Geiselodinium nor Subtilisphaera genera. These two taxa present one apical horn and either one eccentrically located antapical horn, or two symmetrically located but unequal horns. The most closely satisfactory genus is Corculodiniunz Batten & Lister, 1988 emend. herein.