The organic matter of lithologic Units V and VI consists ~50%-80% of well-defined plant debris, such as wood fragments of various sizes, fern leaves, fern sporangia, seeds, and cuticles (Table T1, and can thus be classified as structured terrestrial palynodebris. About 20%-40% of the organic debris is finely shredded.
In some samples, especially in Samples 183-1138A-72R-3, 17-20 cm, and 72R-3, 20-25 cm, many resin pieces were found. Usually, these unrounded fragments are not >1 mm in diameter, are translucent or opaque, and are whitish to yellowish in color. Several of these pieces were unsuccessfully analyzed by A. Otto for terpenoid biomarkers, which are typical for conifers.
Most of these samples contain either abundant or frequent sporomorphs, which are usually well preserved and relatively unaltered. The color of these sporomorphs is dark yellow (e.g., conifer pollen) to light brown (spores). The value on the thermal alteration scale (TAS) is estimated to be 2 to 2.5 (Batten, 1996).
The organic matter of lithologic Unit IV is quite different. In Cores 183-1138A-67R to 69R, well-defined terrestrial material is virtually absent. In Cores 183-1138A-68R and 69R, dark brown to black highly altered small wood chips are observed. Core 183-1138A-67R seems to contain nearly exclusively light yellow finely disseminated marine organic matter <10 µm in size.
Among the mesofossils of Cores 183-1138A-71R through 73R, fern remains are by far the most common. There are axes with fern leaves partly still attached and young, spirally curved fern pinnae, so-called fiddle heads (Pl. P1, fig. 1). The individual leaflets of these fern pinnules show, in part, epidermal structures and stomata (Pl. P1, figs. 5, 6). Hundreds of sporangia of the Schizaeales type were picked, comparable with recent Anemia (Pl. P1, fig. 4). Some of these sporangia seem to yield in situ spores because among the kerogen in strewn slides, spore clusters partly inside or close to sporangia are encountered, in one case certainly of gleicheniaceous origin.
Megaspores have been identified in sieved material as well as among the HF-processed palynoflora. These megaspores are derived from lycophytes, more precisely from various Selaginella species and possibly from hitherto unknown taxa.
Wood remains are very common in sedimentary Unit VI. Their sizes range from <1 mm to 3 cm. Some of the wood particles, mainly charcoalified pieces, show excellent preservation of structural detail. In some of these particles, light-microscopic as well as SEM pictures reveal cell structures, mainly tracheids with pits. Thus, most of these particles seem to be conifer wood.
Several seeds are identified by D.H. Mai (pers. comm., 2001) as conifer seeds (Pl. P2, figs. 1, 2), and a small piece of a twig is very similar in appearance to Microcachrys sp., an extant taxon of Tasmania (Pl. P1, fig. 3).
The palynoflora of Cores 183-1138A-67R through 72R consists of marine and terrestrial elements. The terrestrial elements (fungal remains, macro- and miospores, and freshwater algae) are diverse, more than 80 taxa. In the cored interval from Cores 183-1138A-71R through 73R, terrestrial palynomorphs dominate the palynoassemblages but phytoplankton are very rare, whereas in the interval from Cores 183-1138A-67R to 69R, marine elements are absolutely dominant (see Table T1). They consist mainly of dinoflagellate cysts, a few acritarchs, and foraminiferal linings.
This chapter is focused mainly on paleofloristic aspects, such as the reconstruction of the paleovegetation, but also on biostratigraphy. Taxonomic treatment of the terrestrial sporomorphs is based mainly on the publications of Couper (1953), Dettmann (1963, 1973), Filatoff (1975), Norvick and Burger (1975), Burger (1980, 1993), and Backhouse (1988). A complete list of taxonomic references may be found in these papers.
The dinocyst taxonomy used here generally follows the dinoflagellate index of Williams et al. (1998), which contains the full citations of the original papers, and thus these citations are not repeated here. Taxonomic remarks are added in "Appendix A," "Appendix B," and "Appendix C" provide lists of the sporomorph and dinocyst taxa encountered in this material.
The qualitative and quantitative composition of the spore and pollen flora compares very well with mid-Cretaceous floras known from previous studies on material from the Kerguelen Plateau (Mohr and Gee, 1992a, 1992b), as well as with Australian and Antarctic floras of this time interval.
Various botanical groups, such as mosses (bryophytes s.l.), lycopods (Lycophyta), horsetails (Artrophyta), ferns (Filicophyta), Mesozoic seed ferns, conifers (Coniferophyta) and angiosperms (Anthophyta) are present. The botanical affinities of many of these taxa have been discussed previously by Mohr and Gee (1992a). There, botanical as well as paleobotanical literature of in situ occurrences of certain pollen taxa has been summarized. Here, additional information is provided on taxa that have not been observed from the Kerguelen Plateau previously and on new in situ discoveries of the last decade.
The hornwort, liverwort, and moss spore species are quantitatively minor constituents of the sporomorph flora. Compared to the early Albian flora of Site 750, diversity and number of species per sample is slightly higher.
Aequitriradites spinulosus and Foraminifersporis asymmetricus are considered by Dettmann (1963) as spores that might have been produced by Sphaerocarpaceae and Anthocerotaceae. Triporoletes reticulatus might be related to extant Riella/Riccia (Riellaceae). Several fossil spore specimens, most likely of moss origin, are similar in appearance to extant taxa of Asterella (see Playford and Dettmann, 1996).
The spore taxa Stereisporites antiquasporites and Stereisporites granuloides (Filatoff, 1975) belong, most probably, to the Sphagnaceae.
Trilete as well as monolete spores are preserved. The majority of trilete spores were certainly produced by ferns; the monolete spores also come from ferns.
The fernlike plants are represented by various taxa, which are found only occasionally. Calamospora mesozoica is considered to belong to Equisetites. Densoisporites velatus, which is found regularly at a few percentage abundance in the spore spectra, may be derived from Lycopodiaceae or Selaginellaceae. Species of Retitriletes can be correlated to extant Lycopodiaceae. Camarozonosporites, Foveosporites canalis, and Sestrosporites are also considered to be of lycophyte origin (Mohr and Gee, 1992a).
A megaspore, similar to Balmeisporites holodictyus (Cookson and Dettmann, 1958), is common in some of the samples, especially in 183-1138A-72R-1, 122-124 cm. A similar type of megaspore is known from Aptian to Albian strata of South America (Baldoni and Batten, 1991). Krassilov and Golovneva (1999) observed remains of a heterosporous plant from the Cenomanian of West Siberia. This plant produces megaspores that have lateral pockets filled with monolete microspores. Megaspores with attached microspores of the Perinomonoletes type have been identified (Wilde and Hemsley, 2000) from the terrestrial Barremian of western Germany. In the Kerguelen Plateau specimens, we observe adherent microspores as well. Krassilov and Golovneva (1999) suggest either an affinity to lycopsids or water ferns, even though neither group shows the same biological characteristics. Wilde and Hemsley (2000) consider Selaginellales and/or Isoetales as closely connected to these megaspores. Aratrisporites-type spores (Pl. P3, figs. 3, 4), which are found in Cores 183-1138A-72R to 73R, are considered to be related to Isoetalean plants.
Among the ferns, the taxon Cyathidites is most common. Various fossil and extant taxa might be related, such as Coniopteris, Onychiopsis, and extant Dicksonia (Dicksoniaceae) (Traverse, 1988; van Konijnenburg-van Cittert, 1989). Other spore genera such as Cibotiumspora and Ischyosporites might also belong to the Dicksoniaceae. Spores of Gleicheniaceae (Gleicheniidites sp.) are found also in situ in sporangia. Cyatheaceae also probably grew at this location. Spores of Reticulisporites (sensu Uwins and Batten, 1988) are similar to spores of recent Cyathea and Alsophila (Large and Braggins, 1991). More rarely, spores of Osmundaceae (Baculatisporites), Pteridaceae (Contignisporites), and Schizaeaceae, probably including Lygodium (Impardecispora and Biretisporis) were encountered.
Schizaeaceae were rather diverse. Several members of the Cicatricosisporites-Appendicisporites-Plicatella complex were found in this material. Cicatricosisporites accommodates spores consistent with those of Anemia and Mohria. Appendicisporites and Plicatella are also diagnostic of Anemia-type spores, representing the Anemia oblongifolia, Anemia raddiana, and Anemia phyllitidis types (Dettmann and Clifford, 1992). Ruffordiaspora, defined as spores that are similar to those of the fossil taxon Ruffordia goeppertii, belong certainly also to the Schizaeaceae. The same is probably true for the monolete reticulate spore taxon Microfoveolatosporis fromensis.
Nonconifer pollen are rare in the studied material. A few bisaccate Vitreisporites pollen, known to belong to Caytoniales (Caytonanthus) and Alisporites, considered to belong at least partially in the group of Mesozoic seed ferns (Mohr and Gee, 1992a), were encountered. Cycadopites might belong to cycadophytes, which include, besides true cycads, the Bennettitales. Callialasporites and Balmeiopsis limbata have been identified from taxa of Brachyphyllum; however, the question of whether or not Callialasporites is ultimately of araucariacean or possibly podocarpacaen affinity has not been resolved (Archangelsky, 1994). Corollina is known to belong to the extinct conifer family Cheirolepidiaceae.
Podocarpidites might be related to Podocarpus (Podocarpaceae) and Rugubivesiculites and Lygistepollenites to extant Dacrydium (Dettmann, 1994). Microcachrydites antarcticus is usually matched with extant Microcachrys (Cookson, 1947) and Trichotomosulcites with Trisacocladus (Dettmann, 1994).
Angiosperm pollen generally contribute 1% to 3% of the sporomorph assemblage and comprise monocolpate types, such as two species of the genus Clavatipollenites that are generally considered to be of Chloranthaceous origin. Tricolpate and possibly tricolporoidate pollen are present, partly preserved as tetrads. Among the tricolpate specimens, we found the zonal marker Phimopollenites pannosus, indicative of the Phimopollenites pannosus Zone, correlatable to the late Albian. One pollen species might be related to the taxon Afropollis, common in mid-Cretaceous low latitudes.
The general level of development fits very well with descriptions of early angiosperm pollen from Australia and Antarctica (Burger, 1993; Dettmann, 1973; Dettmann and Thomson, 1987), where monocolpate and tricolpate taxa are dominant. At the present time these taxa are not correlatable with extant angiosperms.
In order to understand the floral development through time, 12 samples from Cores 183-1138A-71R to 73R were quantitatively analyzed. They comprise an interval of ~18 m (672-690 mbsf) with two gaps each of ~7 m between 672-679 and 682-689 mbsf. The 3.50-m interval between these gaps is sampled relatively intensely. Thirteen major pollen/spore categories were identified, and 50 to 150 sporomorphs were counted per sample.
All samples show, in principle, a common pattern (Fig. F3). The majority of all sporomorphs of Cores 183-1138A-71R to 73R consist of fern spores (average = 70%-80%). Moss and lycophyte spores are much less common (2%-3%). Podocarp conifers are usually present, with 5% to 8%, Corollina (Cheirolepidiaceae) with 3%-10%. Angiosperms are present in a typical range of 1%-3%. Mesozoic seed ferns, such as Caytoniales (Vitreisporites) and Corystospermales (Alisporites, in part), are very poorly represented.
During the time interval studied, the pattern did not change significantly. However, in the 3.50-m interval we observe three spikes of Cyathidites spores, which might express some kind of cyclicity. In saccate as well as Corollina pollen we see a similar pattern, and also with Densoisporites velatus, but the counted numbers for these latter groups are statistically insufficient.
Several of the taxa encountered in Cores 183-1138A-71R through 73R are stratigraphic index taxa that have well-defined ranges and were used by various authors, such as Dettmann and Playford (1969), Helby et al. (1987), and Dettmann (1994), for correlation purposes. The following taxa are considered to be time indicative (see Fig. F4): the megaspore Balmeisporites sp. and closely related taxa have been reported from the Barremian through the Cenomanian. Balmeisporites holodictyus has been observed in Australian upper Lower Cretaceous strata; a species that is less sculptured has been found in the lower Upper Cretaceous sediments of Victoria (Cookson and Dettmann, 1958). Clavifera triplex has its first appearance during the mid-Albian. Phimopollenites pannosus has a consistent appearance during the late Albian. Interulobites intraverrucatus and Plicatella distocarinata are slightly younger and are not seen in Southern Australian sections before the uppermost Albian. Balmeiopsis limbata, Callialasporites dampieri, and Klukisporites scaberis made their last appearances during the uppermost Cenomanian. Cicatricosisporites hughesii disappeared in the lower Cenomanian. A thorough discussion of various stratigraphic aspects is given below.
Most of the phytoplankton observed in Sections 183-1138A-67R-1 through 72R-1 consists of dinocysts. In a few samples, such as 183-1138A-68R-3, 123-125 cm, acritarchs, mainly Verihachium sp., are also moderately common.
The cyst assemblages are, except for a few samples, not diverse. The most diverse sample (183-1138A-68R-3, 123-125 cm) contains ~20 taxa. Overall, >40 taxa have been determined (Figs. F4, F5). Since many of the analyzed samples are very poor in abundance, (semi)quantitative counts were not possible.
The most common taxa are Cassiculosphaeridia reticulata, Cribroperidinium edwardsii, Cibroperidinium muderongense, Cyclonephelium compactum, Heterosphaeridium heteracanthum, Odontochitina operculata, Palaeohystrichophora infusorioides, and Spiniferites sp.
Several of the taxa encountered are only present in the lower part of the section, such as Ascodinium and Hapsocysta peridictya (Sample 183-1138A-69R-5, 124-126 cm), which was formerly observed in Albian to Cenomanian strata of Australia (Eisenack and Cookson, 1960) and from Albian strata of the Weddell Sea (Mohr, 1990). Conosphaeridium striatoconum and Chatangiella sp. were seen in the upper part of the section (Samples 183-1138A-67R-2, 118-120 cm, to 68R-3, 123-125 cm).
Several of the observed dinocysts are stratigraphic marker taxa, indicative of a late Albian to Coniacian time interval (Helby et al., 1987), as shown in Figure F6.
Short-ranging taxa are Ascodinium parvum (mid-Albian to late Cenomanian), Disphaeria macropyla (latest Albian to Santonian), Litosphaeridium siphonophorum (mid-Albian to mid-Turonian), Microdinium sp. (early Albian to late Turonian), and Conosphaeridium striatoconum (Coniacian to Santonian). Chlamydophorella ambigua was used by Schiĝler and Wilson (1998) as an index taxon for their C. ambigua Zone, which marks the mid-Coniacian of New Zealand. Glaphyrocysta marlboroughensis, a new species from New Zealandian strata, has been recorded by these authors through the late Coniacian and early Santonian. Acme time intervals of certain taxa, such as Cribroperidinium edwardsii (late Cenomanian to mid-Turonian), Microdinium ornatum, and Palaeohystrichophora infusorioides (both latest Cenomanian to Turonian), also seem to be useful for stratigraphic purposes. A detailed stratigraphic discussion is given below.
The affinities of many of the sporomorphs are well known (see above) and reflect the composition and diversity of late Early Cretaceous southern high-latitude floras. The composition of megafloras of the Antarctic and Australian late Early Cretaceous support our palynological results (Douglas, 1994).
Wood remains have been found and described from various localities at southern high latitudes. Falcon-Lang and Cantrill (2000) discussed the anatomical features of silicified tree trunks and stumps from upper Albian strata of Alexander Island, Antarctic Peninsula, and concluded that four taxa might have been present that seem to be related to extant Podocarpaceae, Araucariaceae, and Taxodiaceae (Athrotaxis).
Francis and Coffin (1992) analyzed wood from Core 120-750B-13W of early Albian age. Their conclusion was that these wood pieces all belong to conifers, more precisely to the form genera Podocarpoxylon or Mesembrioxylon, which partly have features of modern Podocarpaceae. At the Kerguelen Plateau, we found, besides conifer wood, which was difficult to determine more precisely, conifer seeds and twigs (probably of Microcachrys). These remains seem to be related mostly to podocarps, but a relationship to Araucariaceae or other conifer families cannot be ruled out completely. The well-studied late Aptian to early Albian southern Australian Koonwarra flora gives additional insight to the understanding of the vegetation of this time period (Drinnan and Chambers, 1986). In the Koonwarra material, Bryophytes, Equisetophytes, and Lycophytes are present and the diversity of ferns is rather high, with 11 taxa. Gymnosperms were also very common with three taxa of Ginkgophytes, Coniferophytes (nine taxa), Mesozoic seed ferns (four taxa), and probably also a few taxa of Gnetophytes. Angiosperm remains were, however, still extremely rare (Taylor and Hickey, 1990).
The studied pollen spectra fit best with the general picture of the early Albian palynoflora from the Kerguelen Plateau (Mohr and Gee, 1992a) but equally reflect differences resulting from the late Albian age of the flora discussed in this paper. By ~100 Ma, angiosperms had become more diverse and more common, whereas Mesozoic seed ferns such as the Caytoniales seem to have diminished as a percentage of the vegetational cover.
The quantitative sporomorph data seem to support the following conclusions about Kerguelen Plateau vegetation and floral change through time: probably a large part of the canopy consisted of podocarp and other conifers. The understory and especially the ground cover was mostly made up of ferns of various growth forms, including tree ferns (e.g., Dicksoniaceae), climbing ferns (Lygodium), and ground species. Angiosperms were certainly present, possibly severely under-represented in the pollen record. Mesozoic seed ferns seem to have been remarkably rare, perhaps diminished by more competitive early angiosperm taxa. On swampy ground, Isoetalean plants might have covered larger areas, as Isoetes does today.
Cores 183-1138A-71R through 73R contain mostly sporomorphs and very few dinocyst specimens that are not time indicative. Therefore, stratigraphic conclusions are based solely on sporomorph taxa. Cores 183-1138A-67R through 69R, however, contain nearly exclusively dinocysts, which are here used for stratigraphic purposes. We follow the widely applied stratigraphic zonation schemes for palynomorphs and dinocysts based on the works of Helby et al. (1987) and their definitions of superzones and zones to correlate the Kerguelen Plateau sections with the Australian Mesozoic (see Fig. F4). We also list the ranges of the stratigraphic index taxa, of which several are found in our material, mostly according to their data (see Fig. F4). Additional data used are from Stover et al. (1996) and Burger (1988).
The spores and pollen encountered at Cores 183-1138A-71R through 73R directly overlying the basaltic basement (radiometric age = 103-95 Ma) are, generally speaking, well known from the Australian late early to early late Cretaceous, and most of these sporomorphs have already been described from the KP (Mohr and Gee, 1992a, 1992b). Several of the taxa encountered in this material are stratigraphic index taxa that have well-defined ranges and were used by various authors, such as Helby et al. (1987) for correlation purposes. According to Helby et al.'s (1987) miospore zonation scheme, the Kerguelen Plateau assemblage is part of the Hoegisporis Superzone, which ranges from mid-Albian to early Turonian in age. To pinpoint the age of the section more precisely, several spore and pollen taxa proved to be useful, such as the cicatricose spores Cicatricosisporites hughesii and Plicatella distocarinata. Their ranges overlap in the upper part of the Phimopollenites pannosus Zone and the lowermost part of the Plicatella distocarinata Zone (formerly Appendicisporites distocarinatus Zone), which indicates a late Albian to earliest Cenomanian age. This age estimate is in good agreement with the presence of a taxon of Balmeisporites, which was morphologically similar to the two species, B. holodictyus and B. glenelgensis. The first species is common in upper Lower Cretaceous strata in South Australia and Victoria, the latter in Cenomanian strata of Victoria.
Another criterion to evaluate the age of the studied section is the evolutionary stage of the angiosperm pollen. In contrast to the lower Albian section of Site 750, where only one of the most basal angiosperm pollen types, Clavatipollenites, has been reported, the material from Hole 1138A yields a variety of angiosperm pollen, about a half-dozen taxa. Among these the tricolpate genus Phimopollenites is characteristic of the Phimopollenites pannosus Zone (late Albian) (Burger, 1993). Younger sediments of Cenomanian to Coniacian age recovered at Site 747 have an even larger diversity of more than a dozen taxa, of which several are already clearly tricolporate, a developmental state that evolved only later (Mohr, 1995). Thus, the flora recovered during Leg 183 is of an intermediate age, most probably late Albian, according to the angiosperm development.
We conclude that the age of the section studied can be only slightly younger than the ages given for the volcanic basement, probably very latest Albian.
The overall composition is very close to dinocyst assemblages described previously from the Kerguelen Plateau, Site 748 (Mohr and Gee, 1992b), of late Cenomanian to early Santonian age, and from the Australian Perth Basin (Cookson and Eisenack, 1974, 1982) Gingin Brook borehole (404-414 ft), which was considered to be of mid-Cretaceous age, and the Balcatta borehole, of possible Albian to Cenomanian age.
When Helby et al.'s (1987) zonation scheme is applied to the Kerguelen Plateau material of Cores 183-1138A-67R through 69R, the dinocyst assemblages fit well with the Heterosphaeridium Superzone because the bulk of our species have their occurrences in this interval (see Fig. F4), which is considered to be of late Albian to Coniacian age.
Several of Helby et al.'s (1987) index taxa, such as Cribroperidinium edwardsii, Microdinium ornatum, and Palaeohystrichosphora infusorioides, all recorded in Cores 183-1138A-68R through 69R, and very common in the Palaeohystrichophora infusorioides Zone, which equates to the latest Cenomanian to late Turonian. Thus, we conclude that most of the samples from Cores 183-1138A-68R through 69R are of this age.
First appearance datums and last appearance datums of a few taxa, recorded in Cores 183-1138A-67R through 68R allow an even more precise dating of the section. The presence of Ascodinium parvum and Canninginopsis cf. denticulata, together with Isabelidinium glabrum in Sample 183-1138A-69R-5, 124-126 cm, makes it probable that this part of the core is still of Cenomanian age.
Litosphaeridium siphonophorum is recorded worldwide from the mid-Albian to the mid-Turonian (Stover et al., 1996). The last occurrence in the Kerguelen Plateau cores is in Sample 183-1138A-68R-2, 124-126 cm. This means that this sample must be mid-Turonian or older. This is also in accordance with the observation that Cribroperidinium muderongense and C. edwardsii are seen often in Sample 183-1138A-68R-3, 57-59 cm, and in Sample 183-1138A-68R-2, 124-126 cm, but only rarely in Sample 183-1138A-67R-2, 118-120 cm. Cribroperidinium edwardsii has its peak occurrence in the Diconodinium multispinum and Palaeohystrichophora infusorioides Zones, which correlate with the mid-Cenomanian to earliest Coniacian, and is only rarely seen later. Kleithriasphaeridium tubulosum (formerly Conosphaeridium tubulosum) in Sample 183-1138A-67R-2, 118-120 cm, might indicate that these strata belong to the Conosphaeridium striatoconus Zone, which is considered to represent the Coniacian. More precisely, the sample might belong to the Chlamydophorella ambigua Zone of Schiĝler and Wilson (1998), which is correlatable to the mid-Coniacian (see above). This might explain why common taxa such as Odontochitina porifera and a variety of Chatangiella species, which are usually common during the late Coniacian and Santonian, have not yet been seen.
These age estimates fit very well with those gathered for other microfossil groups. According to calcareous nannofossils, the cored interval from Cores 183-1138A-67R through 69R might belong to the (mid) Coniacian to the early Turonian, perhaps latest Cenomanian. Core catcher Sample 183-1138A-67R-CC might be of mid to late Turonian age (S.W. Wise, pers. comm., 2000), whereas Sample 183-1138A-67R-2, 34-35 cm, might be of mid-Coniacian age (Coffin, Frey, Wallace, et al., 1999). Cores from Sections 183-1138A-69R-3 and 69R-4 probably represent the mid and early Turonian. The planktonic foraminifers indicate similar ages (Coffin, Frey, Wallace, et al., 1999). Samples 183-1138A-67R-4, 68-71 cm, to 68R-4, 5-7 cm, seem to belong to the Turonian Whiteinella baltica Zone. Sample 183-1138A-69R-5, 81-83 cm, belongs possibly to the upper part of the Praeglobotruncana spp. Zone and might represent the early Turonian.