- © 2004 by AASP Foundation
The Upper Cenomanian ‘black shales’ at Aksudere in Crimea contain a distinct palynological association dominated by prasinophyte phycomata, mainly leiospheres and tasmanitids, and peridinioid dinoflagellate cysts, mainly Trithyrodinium suspectum ukrainense subsp. nov. Prasinophyte dominated assemblages have not previously been reported from Cenomanian–Turonian boundary organic rich deposits. Their presence confirms probable deposition under oxygen-deficient bottom waters in this area.
Organic-rich, dark-coloured mudrocks (‘black shales’) are widespread around the Cenomanian–Turonian boundary level in the Crimea and Caucasus (Naidin, 1981). Their relationship to the contemporary worldwide black shale depositional episode (‘oceanic anoxic event’ of Schlanger and Jenkyns, 1976), has recently been explored by Kopaevich and Walaszczyk (1990), Naidin (1993, 1996) and Naidin and Kiyashko (1994a, 1994b). These investigations were focussed on the reference section at Aksudere, near the town of Bakhchisarai in the Crimean mountain range (Text-Figure 1⇓). At Aksudere, a 70 centimetre thick succession of interbedded black shales, beds 4–8 of Naidin (1993), is developed in paler coloured carbonate deposits (Text-Figure 1⇓).
Kopaevich and Walaszczyk (1990) described beds 4–8 as “dark marly clays with silt-size quartz, glauconite and volcanic material and almost a complete lack of both macro- and microfaunal calcitic remains.” The paucity of calcitic/aragonitic fossils has rendered accurate dating of the beds difficult. Splits of 21 samples (91- series, Text-Figure 1⇑) that were analysed for their geochemistry by Naidin (1993) and Naidin and Kiyashko (1994a, 1994b) are being investigated for different microfossil groups and an integrated report is in progress. This paper summarises the palynomorph distribution, biostratigraphy and inferred palaeoenvironments of the succession and describes a new peridinioid dinoflagellate cyst subspecies, Trithyrodinium suspectum ukrainense that occurs abundantly in the samples.
Weighed, dried rock samples were dissolved in hydrochloric acid (HCl 35%) and hydrofluoric acid (HF 40%) in order to remove carbonate and silicate minerals respectively. Preparations were sieved with 10 μm mesh. Kerogens from many of the paler coloured carbonate lithologies are dominated by phytoplankton and transparent, ‘cloudy’ amorphous organic matter. The latter was removed by a ‘nitric wash’, i.e., two minutes of oxidation with nitric acid (38% HNO3). The black shale kerogens are dominated by amorphous organic matter. Clumps of this material can outnumber palynomorphs at a ratio of several hundreds or thousands to one. Between 4 minutes and 3 hours of oxidation with Schulze’s solution (38% HNO3 supersaturated with potassium chlorate, KClO3), followed by one subsequent rinse with 2% potassium hydroxide (KOH) solution, were used to liberate palynomorphs from the amorphous organic matter (Text-Figure 2⇓). The duration of treatment in Schulze’s solution was determined by organic carbon content, samples with the largest amounts requiring longer oxidations to liberate palynomorphs (Text-Figure 2⇓). Fungal material, modern field contamination, is present in many kerogen samples. All preparations were treated with Bismark Brown stain. Approximately 300 palynomorphs were counted per sample (Table 1⇓). Up to four microscope slides, labelled A–D, were carefully scanned for additional taxa. Several thousand specimens are present on most slides. However, some carbonate samples, notably from beds 1, 3 and 11–23 are relatively low yielding (less than 30 palynomorphs per gam). Full counts are given if less than 300 palynomorphs were encountered. The method used for calculating absolute abundance (i.e., the estimated number of palynomorphs per gram of dried sediment) is described elsewhere (Dodsworth, 2000). All microscope slides used in this study are curated at the Palynology Research Facility, University of Sheffield.
Division DINOFLAGELLATA (Butschli 1885)
Fensome et al. 1993
Class DINOPHYCEAE Pascher 1914
Order GONYAULACALES Taylor 1980
Suborder PERIDINIINEAE Fott 1959
Family DEFLANDREACEAE Eisenack 1954
Genus Trithyrodinium Drugg 1967
Trithyrodinium suspectum ukrainense subsp. nov.
Plate 1, figs. 1–8⇓
Designation of holotype.
Plate 1, figs. 1 & 7⇓. Slide 9104-1(A) W48/2. Specimen dimensions: 49 μm long x 43 μm wide. The slide is curated in the Palynology Research Facility, University of Sheffield. The collection number for the type is ML5786.
Designation of paratypes.
The taxon occurs in Aksudere samples 9101-1 to 9123. The stratigraphic range is Upper Cenomanian to ?Lower Turonian.
Derivation of name.
The taxon is named after the country Ukraine, in which it is first reported here.
The taxon is a relatively small, thin walled and sparsely ornamented subspecies of Trithyrodinium suspectum. The endocyst is c. 0.5–0.75 μm thick and is ornamented with small (< 0.5–1 μm) grana/pilae that are isolated and evenly scattered or, to an extent, locally clustered. Areas devoid of any obvious ornament, c. 1–5 μm across are characteristic.
The endophragm is subspherical to ellipsoidal in shape. The periphragm is subspherical to ellipsoidal with a short apical horn and two poorly developed antapical horns of nearly equal size, although the right antapical horn is usually slightly longer than the left one. The cyst is circumcavate. The periphragm may be partly or completely missing in some specimens. The endophragm is ornamented with grana. Grana are generally less than 1μ min size but may develop into pilae up to 2 μm in length. The grana/pilae are isolated and evenly scattered, as seen on the paratype (Plate 1, fig. 8⇓) or, to an extent, locally clustered in atabular patches, up to 15 μm in diameter, as seen on the holotype (Plate 1, fig. 7⇓). Areas devoid of any obvious ornament, c.1–5 μm across, occur between grana/pilae. The periphragm is generally smooth but may possess sparse grana. Paratabulation is usually indicated by the archaeopyle(s) only. The archaeopyle is intercalary, type 3I/3I. The periarchaeopyle is rarely discernible. In the endocyst, excystment features are occasionally absent or restricted to sutures around three intercalary plates (e.g. Plate 1, fig. 8⇓). One to three plates are lost in archaeopyle formation with sutures surrounding any of the three plates that remain attached. The cingulum is generally not indicated but on some specimens the position is indicated by folds in the endophragm and periphragm and by shallow local concavities at the lateral margins. The sulcus is not indicated.
The size is intermediate: Pericyst, average length = 54 μm, range 40–72 μm, average width = 46 μm, range 36–62 μm, measured specimens (n) = 52. Endocyst, average length = 45 μm, range 32–68 μm, average width = 42 μm, range 32–60 μm, n = 100.
Trithyrodinium suspectum ukrainense has features in common with Trithyrodinium suspectum sensu stricto and Trithyrodinium evitii. T. suspectum sensu stricto differs in being larger, possessing a thicker endocyst and denser ornament. The type material described by Manum and Cookson (1964) documented specimen length 91–118 μm, width 65–78 μm and wall thickness 2.5–4 μm (n = 4). Davey (1969) assigned smaller, thinner walled specimens to T. suspectum, length 61–68 μm, width 56–59 μm and wall thickness 1.5–2 μm (n = 5). Ioannides (1986) gave measurements for endocyst length 55–75 μm, width 55–72 μm and wall thickness 1.5–3.5 μm. In T. suspectum sensu stricto, the endophragm ornament of grana/rod-like structures is reported by all these authors to be more densely packed, lacking the c.1–5 μm spaces developed between ornament elements seen in T. suspectum ukrainense.
Trithyrodinium evitii and T. suspectum ukrainense are comparable in terms of size and endophragm wall thickness. T. evitii differs in lacking prominent periphragm apical and antapical horns, possessing a smooth to minutely ornamented endophragm that is sometimes covered with a brown organic layer and lacking the distinct grana/pilae of T. suspectum ukrainense.
In carbonate samples taken from below and above the black shales at Aksudere, slightly thicker walls (c. 0.75–1 μm thick, excluding ornament) were recorded from some specimens. These are assigned to T.cf. suspectum ukrainense (Table 1⇑). Many specimens observed from black shales in this study are from preparations that have been oxidised with Schulze’s solution (Text-Fig. 2⇑). This may have resulted in slight bleaching or even thinning of endophragm walls. In a previous work (Schrank, 1988), extended oxidation has been shown to affect the degree of cavation in certain peridinioids. Specimens from unoxidised carbonate preparations in the present study (e.g. Plate 1, fig. 6⇓) are of similar size and exhibit similar cavation and ornament to those from the extended oxidation preparations. Nøhr-Hansen & Dam (1999) demonstrated that oxidation with concentrated nitric acid and subsequent treatment with potassium hydroxide resulted in the removal of a brown non-sporopollenin organic layer that surrounds the endocyst of T. evitii. Layers of this type were not observed on T. suspectum ukrainense endocysts from unoxidised kerogen or extended oxidation preparations.
The specimens present in the Aksudere material are considered to be close in morphology to specimens that are often included in the taxon Trithyrodinium suspectum. They are sufficiently alike and sufficiently different from the type material of T. suspectum, in terms of size, wall thickness and possessing areas devoid of ornament, to warrant the erection of the new subspecies T. suspectum ukrainense.
The distribution of palynomorphs in the Aksudere succession is summarised in Table 1⇑. Over 100 dinoflagellate cyst taxa were recorded. Assemblages are comparable to those reported from Upper Cenomanian–Lower Turonian deposits in western Europe and North America. Beds 2–9 contain taxa including Litosphaeridium siphoniphorum (e.g. Plate 2, fig. 1⇓). Although only one specimen was recorded in the 300 count for samples from beds 2–9 (Table 1⇑), up to 10 specimens were encountered on each slide during scanning of further material. Associated with these frequent occurrences of L. siphoniphorum are rare Adnatosphaeridium tutulosum (e.g. Plate 2, fig 9⇓), Carpodinium obliquicostatum (e.g. Plate 2, fig. 6⇓), Gonyaulacysta cassidata (e.g. Plate 2, fig. 4⇓) and Microdinium setosum (e.g. Plate 2, fig. 7⇓). Comparison with the detailed distribution of range tops of these taxa in Colorado, England, France and Germany (Dodsworth, 1996, 2000), indicates that beds 2–9 are no younger than intra-Late Cenomanian age, equivalent to mid-levels of the Metoicoceras geslinianum ammonite Zone, or stratigraphically lower. There are no known palynological markers for delimiting the stratigraphical position of beds 2–9 in the lower M. geslinianum or older ammonite zones of the Upper Cenomanian. Foraminiferal and inoceramid evidence from the underlying white carbonates (Kopaevich and Walazczyk, 1990) does however constrain the age to Late Cenomanian.
Epelidosphaeridia spinosa (e.g. Plate 2, fig. 11⇓) and Protoellipsodinium spinocristatum (e.g. Plate 2, fig. 12⇓) are present in beds 4–8 (Table 1⇑). E. spinosa has a consistent Middle Cenomanian range top in northwest Europe (Foucher, 1981; Costa and Davey, 1992; Tocher and Jarvis, 1994) though it ranges into the Turonian in the Vocontian Trough, southeast France (Courtinat et al., 1991), in northern Spain (Lamolda and Mao, 1999), Bulgaria (Pavlishina and Minev, 1998) and North America (Haq et al., 1988; Li and Habib, 1996). P. spinocristatum has not previously been recorded from deposits younger than Albian (e.g. Singh, 1971; Davey and Verdier, 1971, 1973; Verdier, 1975). With the exception of isolated specimens of the dinoflagellate cyst Hapsocysta peridictya (reported range Albian–Middle Cenomanian; sample 08-2) and the gymnosperm pollen Callialasporites dampieri (reported range Jurassic–Albian; sample 04-2), there are no obvious associated reworked Albian to Middle Cenomanian taxa in beds 4–8 (Table 1⇑). It is likely that the range tops of in situ E. spinosa and P. spinocristatum are relatively high, i.e., Upper Cenomanian, in the Crimean region.
Beds 6–23 contain frequent/common Cyclonephelium membraniphorum (e.g. Plate 2, fig. 2⇓). This may correlate with an inter-regional influx of the taxon reported from Upper Cenomanian–lowermost Turonian deposits (M. geslinianum/S. gracile–W. coloradoense Ammonite Zones) in other European and North American localities (Dodsworth, 2000, fig. 12). Naidin (1993, 1996) estimated the position of the Cenomanian–Turonian boundary at the top of the black shales (top of bed 8) but did not discuss his criteria for the boundary pick. It is tentatively raised to the top of bed 9 here, i.e., the top of intra-Late Cenomanian marker persistent/frequent Litosphaeridium siphoniphorum, but could be higher. The first foraminiferal and inoceramid evidence for the Lower Turonian was encountered in a carbonate sample collected three and a half metres above the black shales (Kopaevich and Walazczyk, 1990). Lower Turonian dinoflagellate cyst markers such as Heterosphaeridium difficile were not recorded in the present study.
Palynomorph recovery from the black shales of beds 4–8 is mainly composed of prasinophyte phycomata and three dinoflagellate cysts, Trithyrodinium suspectum ukrainense, Palaeohystrichophora infusorioides and Spiniferites ‘group’ (Text-Figure 3⇓). The adjacent organic-poor, paler coloured lithologies are characterised by P. infusorioides and Spiniferites ‘group’, along with Heterosphaeridium? heteracanthum, Odontochitina spp. and, from bed 6 upward, Cyclonephelium membraniphorum (Table 1⇑ and Text-Figure 3⇓). The peridinioid Subtilisphaera spp. is also abundant in assemblages from beds below the black shales (Table 1⇑). Apart from Dapsilidinium ambiguum (e.g. Plate 2, fig. 3⇓), the other taxa present are in general only found in an abundance of less than one percent. Terrigenous palyno-morphs, mainly trilete spores and bisaccate pollen, are generally a minor component of assemblages (Table 1⇑). The palynological assemblages from black shales and carbonates are consistent with a distal marine paleoenvironment.
In Text-Figure 2⇑, total organic carbon (TOC) content, duration of extended oxidation and palynomorph recovery from oxidised preparations are plotted. There is a broad correlation between TOC and palynomorphs per gram. However, in the lowest black shale sample (04-1), exceptionally high palynomorph recovery (c. 20000 per gram) correlates with relatively low TOC (0.5 wt %). It is possible that lower corresponding palynomorph per gram values in the overlying black shale samples reflect an overall reduction in productivity of phycomata/cystforming phytoplankton. However, the duration of extended oxidation was longer in many of these samples (Text-Figure 2⇑). This may have resulted in greater palynomorph loss during processing (see below).
Palynological assemblages from beds 4–8 have been modified, to a largely unmeasurable extent, by laboratory oxidation. Experiments with extended oxidation demonstrate that increasing levels of treatment lead to progressively higher proportions of peridinioids and, to a lesser extent, prasinophytes in resulting preparations (Text-Figure 4⇑). While extended oxidation was necessary on kerogen from the black shales, and facilitated documentation of the diversified dinoflagellate cyst assemblages present, its use clearly precludes detailed statistical treatment of relative abundance data. The oxidation used on the preparations documented here approximately corresponds to the levels represented by oxidation stages 2–3 in Text-Figure 4⇑.
The dominance of prasinophyte phycomata and T. ukrainense in the black shales (Table 1⇑) is partially an artifact of oxidation. However, the absolute abundance data indicate that both are present in large quantities in these lithologies (Text-Figure 3⇑). Large proportions of prasinophytes have not previously been reported from Cenomanian–Turonian black shales. Their dominance, relative to dinoflagellate cysts, in organic-rich strata that were deposited under oxygen-deficient bottom waters, has been documented from other intervals throughout the Mesozoic (e.g. Prauss et al., 1991). Large proportions of peridinioids, i.e., Eurydinium saxoniense, have been reported from Cenomanian–Turonian black shales in northern Germany (Marshall and Batten, 1988). The relative paucity of prasinophytes in adjacent paler coloured lithologies probably reflects cessations in their supply to the sediment during relatively well-oxygenated, open marine conditions.
The main pulse of oxygen-deficient bottom waters/sediments in this area probably occurred during the deposition of beds 4–5 (samples 04-1 to 05); the highest TOC and prasinophytes per gram values are recorded from these levels (Text-Figures 2⇑ & 3⇑). A temporary amelioration appears to have occurred in lower bed 6 times (sample 06-1) with a reduction in prasinophyte numbers and a return to relatively low TOC carbonate lithology. A second pulse of oxygen-deficient bottom waters/sediments probably occurred from upper bed 6 to bed 8 (samples 06-2 to 08-2).
NOTE ADDED IN PROOF
Since the submission of the revised manuscript, nanno-fossil data from Aksudere has come to my attention (Kopaevich and Kuzmicheva, 2002). The Turonian marker taxon Quadrum gartneri has been recorded from Bed 9 upwards at Aksudere and from Bed 10 upwards at a nearby section (Belaya). These data support the tentative positioning of the Cenomanian–Turonian Stage boundary around the base of Bed 10 (Text-Figure 1⇑).
This project was initiated at the Palynology Research Facility, University of Sheffield. I thank Dmitri Naidin and Malcolm Hart for providing the samples, Ted Spinner and Ken Dorning for discussion of the material, Barry Pygott for assistance with photography for Plate 1⇑, Cath Kirwan and Jennie Houlgrave for assistance with Text-Figure 1⇑, David Batten and an anonymous referee for suggesting improvements to the text.