The dinoflagellate cyst genus Polygonifera; emendation and taxonomic stabilization

The taxonomic identity of the dinoflagellate cyst genus Polygonifera Habib, 1972 is confirmed, through the recognition of additional morphological characters during the study of specimens from Early-Middle Cretaceous sediments of India. A revised definition of this genus is presented. Leberidocysta Stover & Evitt, 1978 is considered a subjective junior synonym of Polygonifera and its type species Leberidocysta (ex: Hexagonifera) chlamydata (Cookson & Eisenack) is transferred to Polygonifera. The other species placed in Leberidocysta are provisionally reattributed to other genera, becoming Craspedodinium (ex: Hexagonifera) defloccata (Davey), Craspedodinium (ex: Palaeostomocystis) pergamentacea (Burger), Hexagonifera laticaudata Vozzhennikova and Hexagonifera scabrata Jain & Taugourdeau-Lantz. A new species, Polygonifera eisenackii, is described. Morphological differences between the camocavate genera Polygonifera and Dingodinium are clarified.

Stover & Evitt (1978, p. 107-108), in their general analysis of fossil dinoflagellates, emended the genus Hexagonifera to include only proximate cysts having an intercalary archaeopyle, a character seen only in the type species H. glabra. The character of archaeopyle in Hexagonifera chlamydata was confirmed as apical (ibid., p. 60). This interpretation is fully acceptable to us, for not only do the illustrations of Cookson & Eisenack (1962, pl. 7, figs. 1-3,5-8) clearly suggest an apical archaeopyle in this species but also our examination of a colour transparency of the holotype (No. 123 in the series prepared for sale by Helby & Partridge, 1980) supports their conclusion. A cavate genus, Leberidocysta, was proposed by Stover & Evitt (ibid., for cavate cysts having an apical archaeopyle, Leberidocysta chlamydata (Cookson & Eisenack) being selected as its type species. Stover & Evitt (1978, p. 75) suggested further that Polygonifera might prove to be a junior synonym of Dingodinium Cookson & Eisenack, 1958, noting the similarity in the nature of their cyst cavation (camocavate). They added that the archaeopyle type in Dingodinium and the possible presence of an apical horn in Polygonifera needed to be determined, to test this hypothesis.

MORPHOLOGY OF THE POLYGONZFERA-LEBERZDOCYSTA GROUP
In our study of Early-Middle Cretaceous subsurface sediments of Perivadavadi shallow well-1, Cauvery basin, south India, a large number of well preserved specimens were encountered resembling Polygonifera and the type species of Leberidocysta, L. chlamydata, in shape, nature of cyst cavation and archaeopyle characters. Specimens with operculum detached were most commonly observed, though individuals with an attached operculum were not rare. The archaeopyle, as observed in Polygonifera and in the type species of Leberidocysta, L. chlamydata, is definitely apical.
Preservation of specimens in different orientations results in an apparently variable degree of pericoel development. The frequent folding of the periphragm and its degree of preservation, together with variable cyst flattening, makes the morphology of these dinocysts appear unduly complex. Nevertheless, this provides an excellent opportunity to study their morphological characteristics in greater detail. Our observations are:

Shape
Specimens, when orientated in dorsoventral position, exhibit generally a polygonal to subpolygonal (Figs. 2a,b,3a) or even oval (Figs. la, 3b) periblast and an oval endoblast. In lateral compressions, both periblast and endoblast appear oval (Figs. lb,c,2c). Similar observations were made by Habib (1972, p. 376), when describing the type species P. evittii. No apical horn is seen, though a small "notch" or mamelon may appear to be present on the apical region, due to periphragm folding (Fig. lb).

Cavation
The cyst is camocavate. The apparent degree of pericoel development is greatly affected by periphragm preservation (being delicate, it is often torn or incompletely preserved) and folding. The pericoel may be well developed (Figs. lb,c,2a,b,c,3a,b) or only partially developed (Figs. la, d). In many cases the periphragm may not be preserved at all, or may be apparently developed only in the antapical region (Pl. 2, fig. 6).
Stover & Evitt (1978, p. 75), analysing the genus Polygonifera on the basis of published illustrations of its type species P. evittii, concluded that the periblast remained attached to the endoblast along the dorsal surface only. Our own observations agree with this. However, in view of the absence of either a clear sulcal notch or parasutural features, our interpretation, and also the orientation of specimens illustrated in this study, cannot be considered conclusively established. Regarding the orientation of the specimens of Polygonifera examined in the present study, we believe them to be most commonly preserved either in slightly oblique dorsal (Figs. 2a,b,3a,b) or ventral views (Fig. 2), though lateral compressions are also not uncommon (Figs. l b ,c,d,2c).

Archaeopyle
The archaeopyle is formed by the loss of all the paraplates of the apical series, these being removed as a simple operculum. The number of paraplates appears to be four, as suggested by the occasional partial development of accessory archaeopyle sutures (Figs. l a ; P1. 2, fig. 4). Rarely, only the anterior piece of the operculum on the ventral surface is lost in archaeopyle development, the dorsal section remaining attached (Fig. l fig. 7).
The archaeopyle margin is usually zigzag. Habib (1972, p. 376), studying this feature in P. evittii, considered a sulcal notch and six precingular paraplates to be indicated (ibid., pl. 5, figs. 1,2). Though a sulcal notch could not be unequivocally recognised in any specimen in the present assemblage, its possible position is suggested in a few (see Fig. l

Paratabulation
Parasutural features are usually not well developed. However, one species assigned to this genus does exhibit a fairly full paratabulation. This is Polygonifera (ex: Meiourogonyaulax) stafinensis (Gitmez, 1973, p. 276-278, pl. 3, fig. 2;text-fig. 20) Davey, 1982, p. 31, which reflects the paratabulation: ?', ?Oa, 6", 6c, 6"', 2p, lpv, l"", either by means of elevated crests on a murochorate central body, as interpreted by Gitmez herself (op. cit.) or by crests on the endoblast of a camocavate cyst, as seems to be Davey's interpretation. The latter, more recent conclusion is here accepted as correct for purposes of discussion ; however, Davey's re-interpretations of the morphology of this species is unsupported by illustrations and must be considered unproven, pending a re-examination of the holotype.

Claustrum
A small, regularly or irregularly shaped opening of variable size, corresponding to a claustrum (see Norris, 1978), may be present on the hypocyst. Its position is variable; it may be situated just below the cingulum (Fig. lb) or close to the antapex (Fig. lc). Its irregularity of shape and position casts some doubt upon this being a preformed opening; the possibility of its being the work of a scavenger or predator cannot be excluded.
To overcome present problems in establishing the taxonomic identity of these dinocysts, it is necessary to recognise that the degree of development of the cingulum in Polygonifera is a variable, and not a constant, character, unsuitable as a means for generic separation. In consequence, we consider that the genus Leberidocysta, differentiated only on this character, should not be maintained and that its type species, L . (ex: Hexagonifera) chlamydata, should be transferred to Polygonifera.
The other species attributed to Leberidocysta by Stover & Evitt (1978, p. 60) do not exhibit the morphological characters of Polygonifera ; they are discussed later.

Sulcus
This feature is not usually developed, its position being suggested only by the sulcal notch. Only in Polygonifera stafinensis (Gitmez) is the sulcus relatively readily discernible. However, as noted earlier, we have reservations concerning the allocation of that species to this genus.
General archaeopyle formula: (4a)/(4a); after Evitt (1967), Lentin & Williams (1976) and Norris (1978). Diagnosis. Cyst proximate, camocavate. Periblast polygonal-subpolygonal or oval in ambitus, oval in lateral view; endoblast oval to broadly ellipsoidal. Apical horn lacking; a small notch may develop in the apical region, due to periphragm folding. Periphragm thin, delicate; it may be fully, partially or not at all developed and, when present, is typically folded. Endophragm relatively thicker and typically not distorted. Both endophragm and periphragm are smooth or weakly ornamented (reticulate, foveoreticulate, or finely punctate). No parasutural features are present. Archaeopyle apical, (4a)/(4a) ; peripyle and endopyle are usually distinct. Operculum free or attached. Cingulum indicated by periphragm folding, sometimes distinct and completely developed, in other instances faintly, incompletely or not at all suggested. Typically, no indication of a sulcus is seen other than a poorly marked sulcal notch. A claustrum of variable size and shape may be present in the hypotract.  Remarks. This species is readily distinguishable from P. evittii and P . (ex: Hexagonifera) chlamydata by its laevigate to weakly ornamented endoblast. The endoblast ornamentation in P . evittii is scabrate to granulose and that in P. chlarnydata is verrucose. The degree of development of the cingulum is variable in P. eisenackii, whereas the cingulum is well developed in P . evittii (Habib, 1972, pl. 5, figs. 1,2) and not developed at all in P. chlarnydata (Cookson & Eisenack, 1962, pl. 7, figs. Polygonifera eisenackii exhibits a broad size range. The holotype selected is a complete specimen with attached operculum and measures 89 pm (overall length). The two paratypes (a and b) have been selected to demonstrate the marked variation in size exhibited by this species. They represent extremes in their range, the overall length of paratype a being only 5 5 p m (operculum free) whereas paratype b attains 92 pm. It is conceivable that future studies may favour the subdivision of this species into two subspecies, based mainly on this difference in size. Habib (1972, p. 376) gave the size range of the type species P. evittii as 42-62km, as measured on twelve specimens. It was, however, not indicated whether the specimens measured were complete or with detached operculum. Since the specimens from India, described above, are intermediate to large in size, we prefer to recognise the size range in Polygonifera as intermediate to large, and not small to intermediate as described by Stover & Evitt (1978, p. 75). Other species of Polygonifera : 1-3, 5-8.). (Cookson & Eisenack, 1962, p. 496, pl. 7, figs. 1-3, 5-8) comb. nov., Middle-Late Cretaceous (Albian-Cenomanian). [Formerly Hexagonifera and Leberidocysta.] Polygonifera staffinensis (Gitmez, 1970, p. 276-278, pl. 3, fig. 1 ; text-figs. 20a-b) Davey, 1982, p. 31. Upper Jurassic (Kimmeridgian). [Formerly Meiourogonyaulax. 3

LEBERIDOCYSTA
Leberidocysta defloccata (Davey & Verdier, 1973, p. 198, pl. 3, figs. 6,8) was described as closely resembling Hexagonifera chlarnydata (assigned herein to Polygonifera) in its overall appearance but differing in having a smooth inner body. A comparison of the photographs of the holotype of Davey & Verdier's species (1973, pl. 3, fig. 8) and that ofHexagonifera chlarnydata (Cookson & Eisenack, 1962, pl. 7, fig. 2) reveals marked dissimilarities in their morphological characters. The former has a spherical endoblast and a roughly subspherical periblast, while the latter is characterised by having an oval endoblast and subpolygonal periblast. There is a marked difference in the size of endoblast and periblast in Davey & Verdier's specimen, while this is not the case in Polygonifera (ex: Hexagonifera) chlamydata. Further, whereas Polygonifera chlarnydata is camocavate, this character is not obvious in Hexagonifera defloccata. Recognising these characters, we do not feel it proper to place Davey & Verdier's forms in Polygonifera. The precise placement of this species (previous assignments were to Hexagonifera Davey & Verdier, 1973, p. 198, pl. 3, figs. 6, 8 ; Thalassiphora Lentin & Williams, 1976, p. 85;and Leberidocysta Stover & Evitt, 1978, p. 60) can probably be best determined by re-examination of its type    Vozzhennikova (1967, p. 125-126, pl. 54, fig. 1) and Hexagonifera scabrata Jain & Taugourdeau-Lantz (1973, p. 64, figs. 11,12). In view of poor illustrations and descriptions, which do not enable the determination of archaeopyle character, we are not in a position to suggest anything regarding their precise taxonomic identity ; these species are therefore returned to Hexagonifera.
The species described by Burger, first as Palaeosromocysris pergamentacea (Burger, 1980a, p. 88, pl. 47, figs. 2, 3) and a little later as ?Leberidocysta pergamentacea (Burger, 1980b, p. 272) is morphologically very dissimilar to Polygonifera and cannot be retained in it. For the present, we propose its transfer into Craspedodinium as C . pergarnentacea (Burger), pending re-examination of the type material.

CONCLUSIONS
1) The separate taxonomic identity of Polygonifera is stressed and Leberidocysta is considered to be its subjective junior synonym.
2) Polygonifera and Dingodinium are two very different genera which exhibit similarity in characters of cyst cavation and general sue and shape only (Mehrota & Sarjeant, Ms). The table below summarises their distinguishing morphological characters, in order of importance. 3) The complete and distinctive development of the cingulum in Oxfordian-Kimmeridgian species P. evittii and P. staffinensis (Kimmeridgian), its comparatively weak and less distinct representation in Aptian species P. eisenackii, and the lack of an indication of a cingulum in the Albian-Cenomanian species P. chlamydata seems to be of evolutionary significance. Further studies in this Explanation of Plate 2 All figures are x 960 Polygonifera eisenackii sp. nov.