Preliminary account of Late Devonian palynomorph assemblages from the Bedford Shale and Berea Sandstone Formations of central Ohio, U.S.A.

Diverse assemblages of well preserved miospores and microplankton have been obtained from a series of samples collected across the boundary between the Bedford Shale and Berea Sandstone Formations in central Ohio. The miospores indicate a Late Devonian (pre-Carboniferous Tournaisian)* age for each of the samples. The microplankton support this age determination. A number of microplankton species not previously recorded from North America are present.


INTRODUCTION
Four samples collected across the boundary between the Bedford Shale and Berea Sandstone Formations in central Ohio have yielded well preserved assemblages of both miospores and microplankton. The boundary between these formations has been placed by various authors in either the Upper Devonian or the Mississippian.
De Witt (1970) concluded, on the basis of conodont evidence reported by Haas (1947), that only the basal few feet of the Bedford Shale Formation was of Late Devonian age with the remainder of the Bedford Shale and the entire Berea Sandstone Formation being Early Mississippian (Carboniferous) in age. This conclusion was based on the occurrence in the basal Bedford Shale, of Spathognathodus anteposecornis Scott, a species that is characteristic of the latest Devonian Protognathodus fauna in the Upper Mississippi Valley. Conodonts have not been reported from the upper part of the Bedford Shale or Berea Sandstone Formations. Eames (1978), in contrast, has reported an Upper Devonian age, based on palynological evidence, for the entire Bedford Shale and Berea Sandstone in northern Ohio. Previous palynological work in Ohio by Winslow (1962) and in southern Ontario by McGregor (1967McGregor ( , 1970McGregor ( and 1979 on the Bedford Shale and Berea Sandstone Formations and their lateral equivalents would tend to support this conclusion when direct comparisons are drawn with the well established biostratigraphies in the Ardenno-Rhineland areas of Belgium and West Germany. The samples from Big Walnut Creek, Sunbury, discussed in the present account clearly indicate a Late Devonian (Tn l alower Tn lb) age.
Additional sampling in the Berea Sandstone would be necessary before any attempt could be made to define precisely the Devonian-Carboniferous (Mississippian) boundary.

LITH OSTRAT I GRAPHY
The Bedford Shale and Berea Sandstone Formations have been interpreted as the deposits of an elongate delta system prograding southward across the Ohio Black Shale of the Appalachian Basin during the Late Devonian and Early Carboniferous (Pepper et al., 1954). At its type locality in northeastern Ohio, the Bedford Shale Formation consists of 85ft. of grey and bluish-grey shale with grey, irregularly bedded siltstones at its base. Elsewhere in north-central Ohio, red colouration predominates in the Bedford Shale, but this feature decreases in intensity eastward into Pennsylvania and southward into Kentucky. The sandstone and siltstone contents increase in the same direction. The contact with the underlying Ohio Black Shale is regarded as gradational and conformable throughout most of Ohio, with black shale stringers occurring in the basal part of the Bedford Shale (Pepper et al., 1954, p. 159;Hoover, 1960, p. 27).
The Berea Sandstone consists primarily of fine to medium-grained, pale-coloured, quartz sandstone, very variable in thickness. Generally the lower part of the Berea Sandstone is a massive cross-bedded sandstone interpreted as the infill of channels eroded into the Bedford Shale and in some cases older formations. The upper part of the Berea Sandstone is dominated by thinbedded, oscillation-ripple bedforms thought to represent a transgressive phase of the Bedford-Berea deposition (Pepper et al., 1954). *The definition of the Devonian-Carboniferous boundary proposed by Conil, Groessens & Pirlet (1977) at the top of the cosratus conodont zone, within Tnlb, in Belgium is utilised here. This implies the lowest part of the Tournaisian Series, as previously'defined, is now regarded as uppermost Devonian. The Bedford Shale -Berea Sandstone coqtact is sharp and markedly erosional in northern Ohio, but becomes conformable and gradational to the south, with similar lithological sequences developed in the upper part of the Bedford Shale and the lower part of the Berea Sandstone. In southern Ohio and northern Kentucky, the two formations are thin and cannot be differentiated. At the Sunbury exposure (see Fig. 1) the contact does not appear to be erosional although channelling has been reported in the area (Pepper et al., 1954). The formation boundary is there taken at the upward change from flaggy weathering shaly siltstones and thin sandstones to more massive sandstones with flow rolls.

SAMPLE DETAILS
The Bedford Shale-Berea Sandstone boundary is well exposed in a bluff on the west side of Big Walnut Creek immediately downstream from the Croton Road bridge at Sunbury, Ohio (SE b, SE $, sec. 1, T4N, R17W, Sunbury, 1973,71' quadrangle, Delaware County). Four samples were collected as indicated in Fig. 1 resting on a ripple marked surface. Shale.

STRATIGRAPHIC PALYNOLOGY a) Miospores
The composition of the miospore assemblages from the four samples studied are represented diagrammatically in Table 1. All the assemblages are well preserved, but  The single sample from the base of the Berea Sandstone (sample 1) yielded a miospore population which is closely comparable to those from the underlying Bedford Shale. Representatives of Spelaeotriletes lepidophytus are again quantitatively significant components of the assemblage. Two species, Corbulispora cancellata and Hystricosporites (Dicrospora) porrectus, make their first appearance in this sample and Emphanisporites sp., Hymenozonotriletes explanatus, Vallatisporites pusillites, Rugospora jlexuosa and Calyptosporites triangulatus, present in the Bedford Shale, are absent here. The difference between the assemblages from these two formations may reflect stratigraphical distribution of the components, but may equally be the result of environmental controls.

Comparison with other North American assemblages
The most extensive previous study of late Devonianearly Mississippian palynomorph assemblages in the north-central United States is that of Winslow (1962) based on material from surface and borehole sections in central and northeast Ohio. One of Winslow's sections (Locality 9) was in Franklin County, adjacent to Sunbury, where a comparable Bedford Shale and Berea Sandstone interval was examined together with part of the overlying Sunbury Shale. The results obtained from this section compare closely with those of the present study.  (Winslow, 1962 (Winslow, 1962). Specimen MPK 4012. Sample 3. England   Wicander, 1974. Specimen MPK 401 9. Sample 4. England Finder co-ordinates ordinates N26/3 ( x 500).
England Finder co-ordinates N24/0 ( x 500).  Close comparisons can also be drawn with miospore assemblages documented by Streel & Traverse (1978) in In western Europe, Spelaeotriletes lepidophytus and Vallatisporites pusillites are common components in sediments dated on independent conodont and foraminiferal evidence as Fa 2d -Tn l a in the Dinant Basin of Belgium and northern France. In other areas (i.e. the Namur Basin and Sauerland), both species are known also to extend into the lower part of Tn 1 b. Both species have been utilised as zonal indices with their simultaneous appearance defining the Vallatisporites pusillites -Spelaeotriletes lepidophytus (PL) Miospore Zone, the base of which is close to the base of Fa 2d. Other taxa from this zone in Belgium that were also recorded in the present Ohio samples include Rugospora flexuosa, Retusotriletes incohatus, Auroraspora hyalina and, in the uppermost part of the zone, isolated representatives of Hymenozonotriletes explanatus (see Clayton et al. 1977). In Belgium, the PL Zone is succeeded by the

Dictyotriletes trivialis -Hyrnenozonotriletes explanatus
(TE) Miospore Zone, in which assemblages are characterised by abundant Hyrnenozonotriletes explanatus and Dictyotriletes sp. This zone has, for facies reasons, a limited application and the alternative Verrucosisporites nitidus -Vallatisporites vallatus (NV) Miospore Zone erected in the South Munster Basin of southern Ireland is more widely used. Clayton et al. (1974) have subdivided this zone into two subzones, the lower of which, the Spelaeotriletes lepidophytus -Verrucosisporites nitidus (LN) Subzone, has its base defined by the appearance of Verrucosisporites nitidus, which is present in the Bedford Shale and Berea Sandstone assemblages examined here. This subzone is also characterised by the continued occurrence of Spelaeotriletes lepidophytus and Vallatisporites pusillites (albeit with a lower frequency than in the preceeding PL Zone). Although a precise independent dating of the Irish sequence of miospore zones is not possible, it is probable that the LN Subzone of the NV Miospore Zone is of Late Devonian (lower Tn lb) age. The close similarity with the assemblages from the upper part of the Bedford Shale and the lower part of the Berea Sandstone would appear sufficient justification for a comparable lower Tn l b age assignment.

b) Microplankton
Acritarchs and other microplankton (Prasinophyta) were obtained from each of the four samples. Specimens are generally well preserved, but are not as abundant as the miospores and the species diversity in each sample is low with relatively few taxa present. The microplankton assemblages from all four samples are dominated by species of Gorgonisphaeridiurn. The composition of the assemblages is shown in Table 2. There are no major changes in composition between the assemblages from the Bedford Shale (samples 2, 3 and 4) although slight differences do occur. The assemblage from the Berea Sandstone (sample 1) is very similar in composition to those from the Bedford Shale, but contains even fewer taxa and the microplankton is less abundant. These slight differences between the microplankton assemblages may be important, either stratigraphically or ecologically, but given the relatively sparse data available, it is not possible to comment on any significance at present .   (1 974, 1975) has reported specifically on the microplankton from the Bedford Shale and the underlying Cleveland and Chagrin Shale Members of the Ohio Black Shale. In common with the present study, Wicander noted that levels of microplankton species diversity and abundance are low in the Bedford Shale. Wicander's material was obtained from the Barberton test core 3, drilled in Summit County, northeastern Ohio. Comparison of sample depths given by Wicander (1975) with stratigraphical data presented by Winslow  (1 962) indicates that Wicander's uppermost sample was collected immediately below the contact of the Bedford Shale with the Berea Sandstone, a position equivalent to that of sample 2 in this study. Wicander (1975) recorded only six taxa in his uppermost sample, five of which are has not been recorded in this study, but the genus Pterospermella is represented in both the Bedford Shale and Berea Sandstone. Another species present in sample 2, Unellium elongatum, was also described by Wicander from the Bedford Shale, but he recorded its latest occurrence within the top loft. of the Bedford Shale, not immediately below the Berea Sandstone as in this account. The occurrence of all these taxa in samples 3 and/or 4 is however consistent with Wicander's record of their occurrence in the Bedford Shale.
The major difference between Wicander's study and the present account lies in the occurrence of species not previously recorded from the Bedford Shale or Berea Sandstone. However, four species do have North American records. Stellinium comptum, Stellinium micropolygonale (as Stellinium octoaster) and Puteoscortum polyankistrum have been recorded from the Antrim Shale of FrasnianIFamennian age in Indiana (Wicander & Loeblich, 1977). Stellinium micropolygonale has been recorded from the Upper Devonian of Alberta (Staplin, 1961, as Veryhachium octoaster), from the Late Devonian Chagrin and Cleveland Shales underlying the Bedford Shale in Ohio (Wicander, 1974(Wicander, ,1975 as Stellinium octoaster), and from the Emsian, Eifelian and Givetian of the Moose River Basin, Ontario (Playford, 1977). The present record of these species in the Bedford Shale and Berea Sandstone is later than their previous North American records, and implies either reworking or an upward extension of their ranges. In the case of Stellinium micropolygonale, an upward extension of its range is quite possible since the species has been recorded from the latest Devonian elsewhere (e.g. Lanzoni & Magloire, 1969;Playford, 1976 and see the discussion below). The fourth species with a previous North American record, Gorgonisphaeridium winslowii, was recorded from the Lower Mississippian Banff Formation of Alberta (Staplin et al., 1965). Lower Tournaisian (i.e. Carboniferous), Gorgonisphaeridium ohioense (recorded as "Baltisphaeridium sp. No. 419-47") occurs in the Upper Famennian and Strunian, and Stellinium micropolygonale ( = Veryhachium octoaster) and Umbellasphaeridium saharicum (recorded as "acritarche a entonnoirs' no. 441-33") are restricted to the Strunian. Subsequent work in the Algerian Sahara has extended the range of some of these species. Jardine et al. (1 974) recorded Duvernaysphaera radiata between the Couvinian or Givetian and the Strunian, Stellinium micropolygonale ( = Stellinium octoaster) between the Siegenian and the Strunian, Maranhites mosesii between the Frasnian and the Strunian and Umbellasphaeridium saharicum between the Frasnian or Famennian and the Strunian. Attaret al. (1980) confirmed that Maranhites mosesii is present in the Lower Tournaisian of the Basin d'Illizi, also in the Algerian Sahara, whereas Umbellasphaeridium saharicum does not occur above the Strunian. Attar et al. also illustrated a specimen of Gorgonisphaeridium winslowii from the Strunian.

CONCLUSIONS
The miospores recorded in this account from the Bedford Shale and Berea Sandstone provide the principal evidence for a Late Devonian (pre-Carboniferous Tournaisian) age and most of the microplankton recorded are consistent with this. Gorgonisphaeridium ohioense, Stellinium micropolygonale and Umbellasphaeridium saharicum do not occur above the Late Devonian in sections which cross the Devonian -Carboniferous boundary in North Africa (Lanzoni & Magloire, 1969;Attar et al., 1980) and in Western Australia (Playford, 1976). The ranges of other species present, as documented above, coincide in part with the latest Devonian. Stellinium comptum and Puteoscortum polyankistrum are the exceptions, and have not been recorded previously above the Famennian.
The ages proposed herein and by Eames (1978) for the Bedford Shale and Berea Sandstone in central and northeastern Ohio mean that the published ranges of several species need to be revised. These are the species for which a Mississippian occurrence is based solely on Wicander's (1974Wicander's ( ,1975 record of their presence in the Bedford Shale which he regarded as being of Mississippian age following Winslow's (1962) account. They include Exochoderma irregulare, Gorgonisphaeridium ohioense, Veryhachium roscidum and Unellium elongatum. There is, so far, no evidence that any of these or the other species described by Wicander range upwards into the Carboniferous. Wicander & Loeblich (1977) have commented on the apparent rarity of cosmopolitan microplankton species in the Late Devonian, pointing out that only five species occurring in the Late Devonian of North America had been recorded elsewhere in the world. However, Vanguestaine (1 978) and Martin (1 98 1) have recorded a number of North American species in the Late Devonian of Belgium, and the present work documents the occurrence in North America of species previously recorded from North Africa, South America and Australia. There are still considerable differences between worldwide Late Devonian microplankton assemblages. A number of species described from North America have yet to be recorded elsewhere, but some taxa appear to be more widespread than hitherto realised. Further work is needed to clarify the extent of microplankton provincialism in the late Devonian.
Drs. Molyneux and Owens publish with the permission of the Director of the British Geological Survey (N .E. R .C .) .