Numerous palynological assemblages from Late Cretaceous-Neogene deepwater to shallow-marine sediments have been recovered from the continental margins of Antarctica, from adjacent elevated seafloor areas such as the Kerguelen Plateau and South Tasman Rise in the Southern Ocean, and from the Falkland Plateau in the Southwest Atlantic (Webb, 1990; Truswell, 1997).
Many taxa are found recycled (Truswell, 1983; Truswell and Drewry, 1984), but sections likely to include in situ assemblages range in age from Paleozoic to late Quaternary. Key references are Permian (Kemp et al., 1977; Dibner, 1975; Playford, 1990; Farabee et al., 1991; Francis et al., 1993; Lindstrom, 1994, 1995a, 1995b; Larson et al., 1990; Askin, 1997), Triassic (Farabee et al., 1989; Foster et al., 1994), Jurassic-Cretaceous (Askin, 1983, 1989; Domack et al., 1980; Dettmann and Thompson, 1987; Dettmann, 1989; Askin, 1990a, 1990b; Askin et al., 1991; Duane, 1994; Truswell et al., 1999), Paleogene (Cranwell et al., 1960; McIntyre and Wilson, 1966; Wilson, 1967a, 1975; Haskell and Wilson, 1975; Harris, 1976; Hall, 1977; Brady and Martin, 1979; Goodman and Ford, 1983; Mildenhall, 1989; Wrenn and Hart, 1988; Askin, 1990a, 1990b, 2000; Askin et al., 1991; Truswell, 1990, 1991, 1997; Harris et al., 1996; Levy and Harwood, 2000), and Neogene-Quaternary (Mildenhall, 1989; Harris et al., 1996; Raine, 1998). A review of the paleobotanical evidence for angiosperm-dominated plant communities in Antarctica during the Cretaceous and Tertiary was presented by Hill and Scriven (1995). An overview of the Paleocene-Eocene biostratigraphy (including palynology) of the Mac.Robertson Shelf and the western sector of Prydz Bay was provided by Quilty et al. (1999).
Data from the above works provide a regional chronostratigraphy and phytogeographic framework against which the Prydz Bay microfloras can be assessed in three contexts:
This dinocyst flora was first identified from glacial erratics in the Ross Sea region (Wilson, 1967a, 1967b). The distinctive high-latitude suite is centered around occurrences of the late early Eocene-Oligocene species A. antarcticum. The flora tends to be dominated by morphologically variable species of Deflandrea, which historically have been grouped around broadly defined taxa such as D. antarctica and related species such as Deflandrea obeisfieldensis. The Prydz Bay species informally named D. "prydzensis" is part of the same complex.
Although knowledge of the geographic and, more importantly, stratigraphic distribution of this flora remains uncertain (see reviews by Wrenn and Hart [1988] and Truswell [1997]), recent reevaluations of both the marine and terrestrial microfloras from the McMurdo glacial erratics have contributed toward understanding the age constraints. Levy and Harwood (2000) reviewed biostratigraphic controls by comparing species ranges for dinocysts with calcareous nannofossil datums from southern high-latitude sites, including Deep Sea Drilling Project (DSDP) sites toward the South Tasman Rise (Sites 280, 281, 282, and 283) and in the Weddell Sea (Site 696B) and South Atlantic (Sites 511 and 513). They noted that among the McMurdo erratics, the best-known, and relatively diverse, dinocyst assemblages appear to be confined to the middle and late Eocene; assemblages considered to be early Oligocene in age are very restricted in their diversity.
It is becoming increasingly apparent that the flora is subject to geographic variation in terms of taxonomic differentiation and species composition. For example, confusion surrounds the taxonomic limits of morphotypes within the broadly defined D. antarctica complex and assemblages from East Antarctica appear to be generally less diverse than those from the Ross and Weddell Sea regions.
From four sites drilled across the center of Prydz Bay during Leg 119, Truswell (1991) identified three major groups of palynomorphs, namely, Permian bisaccate gymnosperms, long-ranging Cretaceous-Paleogene spores and pollen, and Eocene-early Oligocene dinocysts.
Site 742, drilled in 410 m of water in the middle of Prydz Bay, yielded essentially the same assemblage of dinocysts as that from 142.5-220.85 mbsf at Site 1166, including diverse Nothofagidites, Deflandrea spp., E. partridgei, and Vozzhenikovia apertura. Recovery was, however, sparse; recycled Permian forms were common, and Truswell (1991) was careful to point out that recycling of the whole assemblage remains a possibility.
At the same site, 742A, Core 34 yielded what were described as nonmarine floras of Late Cretaceous age that may be correlative with the Late Cretaceous at Site 1166. Abundant spores and pollen were recovered, including P. mawsonii, Cicatricosisporites australiensis, C. bullatus, M. antarcticus, Kraeuselisporites majus, Podocarpididites spp., and Retitriletes spp. An age of Turonian or younger was suggested on the basis of the miospore evidence (Truswell, 1991), but further processing and re-examination is required to provide a more precise correlation with Site 1166.
Eocene dinocyst assemblages similar to those from Site 1166 are preserved in samples taken from elongate troughs running at right angles to the coast across Mac.Robertson Shelf, west of Prydz Bay, although the diversity is higher and the sequences appear to be slightly older (see Quilty et al., 1999). For example, Site 149/GC47, at the seaward end of Iceberg Alley, contains Tritonites pandus, a species that is restricted to late middle Eocene sediments in the Gippsland Basin. The same species is present in samples from site GC10 from the wall of the Neilsen Basin, a depression that is floored by Late Jurassic-Early Cretaceous sediments (see Truswell et al., 1999).
The "Transantarctic Flora" has been recorded from the glacial erratics at Black Island and Minna Bluff in McMurdo Sound (see above and Cranwell et al., 1960; McIntyre and Wilson, 1966; Wilson, 1967a, 1967b; Askin, 2000; Levy and Harwood, 2000). It has also been reported from suggested Oligocene and Neogene intervals drilled off the Victoria Land coast in the MSSTS-1, CIROS-1, and CRP-1 core holes (Truswell, 1986; Wilson, 1989; Mildenhall, 1989; Raine, 1998).
Among sites drilled in the Ross Sea region during DSDP Leg 28 (1972-1973), Site 270 yielded abundant palynomorphs (Kemp, 1975; Kemp and Barrett, 1975). At that site, an apparent glaciomarine sequence, lithologically described as a pebbly silty claystone, overlies glauconites dated as 26 Ma. Sediments at the base of the glaciomarine sequence (Subunit 2J) are rich in pollen and spores and also contain much degraded plant tissue. The spore and pollen assemblage is dominated by Nothofagidites; there are four or five species of Proteaceae, some Myrtaceae, Podocarpus, and rare cyatheaceous fern spores. This assemblage was suggested to be in situ (Kemp and Barrett, 1975) and to represent a coastal vegetation that persisted into the region into the late Oligocene. However, all palynomorphs identified within it are long-ranging, so the possibility remains that this suite was recycled from older sediments during the Oligocene (see Truswell, 1990).
Of particular significance to the dating and ecological interpretation of the Prydz Bay microfloras is CIROS-1, which records two periods of sedimentation on the margin of an intracontinental rift basin—from latest Eocene or earliest Oligocene (702-366 mbsf) and early late Oligocene-early Miocene (360-0 mbsf) (see Harwood et al, 1989).
Importantly, striated pebbles are found throughout the core, although occurrences are less common near the base (Hambrey et al., 1989). These show that glaciers accumulating on the tectonically rising Transantarctic Mountains extended to the coast by earliest Oligocene time.
The composition of dinocyst floras recovered between 697 and 473 mbsf in CIROS-l is closely similar to those at the Prydz Bay Site 1166 (except for the absence of the manuscript species D. "prydzensis" and the presence of A. antarcticum). The assemblages are dominated by V. apertura and include Alterbidinium (Deflandrea) asymmetricum, A. antarcticum, D. antarctica, E. partridgei, Hystrichosphaeridium tubiferum (Pl. P3, fig. 4), Lejeunocystia sp., Spinidinium macmurdoense, Turbiosphaera filosa, and Vozzhenikovia rotundum.
Additional work is required to determine whether other less distinctive dinocysts found in CIROS-1 are present at Site 1166 (e.g., Deflandrea granulata, Phthanoperidium spp., and Tectatodinium cf. pellitum). Dinocyst floras above 473 mbsf are depauperate and lack D. antarctica, E. partridgei, H. tubiferum, Lejeunocysta sp., and T. filosa but include A. asymmetricum, S. macmurdoense, and V. rotundum.
If similar time distributions apply in Prydz Bay, then the minimum age of samples at 142.5-148.36 mbsf is early early Oligocene, irrespective of known later occurrences of E. partridgei.
Species composition and relative abundance of spores and pollen in the CIROS-1 sequence also closely resemble the Site 1166 microflora (see plates 1-3 in Mildenhall, 1989). For example, the assemblages include essentially the same recycled Permian elements, whereas the component believed to be in situ is dominated by species belonging to the N. flemingii and N. lachlaniae complexes.
Assuming that trace records of Asteraceae and Myrtaceae are contaminants, then the chief differences between the Prydz Bay and Ross Sea microfloral sequences are (1) the Ross Sea lacks, at present, any evidence for a diverse Late Cretaceous suite as found at Site 1166; (2) the CIROS-1 Tertiary suite contains Casuarinaceae (which has also been reported from MSSTS-1, CRP-1, and as a recycled element in surficial Ross Sea sediments); and (3) the CIROS-1 assemblages include two taxa (Chenopodipollis chenopodiaceoides and Corsinipollis epilobioides) that first appear in the early Oligocene in southern Australia (see Macphail and Truswell, 1989). It is noted that it is difficult to distinguish subrecent C. chenopodiaceoides from modern Chenopodiaceae pollen contaminants (present in the Prydz Bay samples), whereas C. epilobioides first appears as early as the middle-late Eocene in New Zealand (D. Pocknall, cited in Mildenhall, 1989).
Another form reported from CIROS-1 that was not observed in Prydz Bay is a tricolpate grain referred to as Perfotricolpites digitatus. This is approximately one-half the size of specimens from southeastern Australia (see Macphail, 1999). The presence of P. digitatus in Antarctica is ecologically anomalous, since (1) the closest living equivalent is pollen produced by the now wholly tropical genus Merrimia (Convulvulaceae) and (2) although P. digitatus first appears in the Murray Basin in the late Eocene, the species has not been recorded eastward in the Gippsland Basin or in Tasmania. The same form, however, was reported from Sirius Group sediments in the Transantarctic Mountains (Askin and Markgraf, 1986), where it occurs within a Nothofagus community. There, affinities with Labiatae or Polygonaceae were suggested. The species has also been figured as Tricolpites sp. A in lower Miocene sediments penetrated during the drilling of the Cape Roberts (CRP-1) borehole in the Ross Sea (Raine, 1998, fig. 2a) and occurs too as a recycled element in Ross Sea sediments (see Truswell, 1983, pl. 3, figs. 23, 24).
Numerous modern examples exist of rainforest growing adjacent to coastal glaciers, for example, in Tierra del Fuego. Mildenhall (1989) argued that the CIROS-1 microfloras represent gymnosperm Nothofagus rainforest surviving in coastal refugia, not rainforest scrublands or scrub. His reasons included the high diversity of Nothofagus relative to existing cold-climate shrublands in South America. The presence in the core of Nothofagus pollen clumps, representing whole anthers and therefore local sources, suggests nearby vegetation. Also recovered was a leaf related to N. gunnii, a winter-deciduous small to tall shrub growing in subalpine to alpine habitats in Tasmania. This interpretation can be challenged on the grounds that pollen does not indicate plant habit per se (Nothofagus species flower profusely irrespective of size), whereas N. gunnii can extend below the timberline along cold-air drainage lines.
Drilling from annual sea ice at Cape Roberts penetrated Tertiary sequences considered to range in age from early Miocene to late Eocene with biostratigraphically useful palynomorph assemblages being recovered from early Oligocene and early Miocene intervals (Raine, 1998; Askin and Raine, 2000; Raine and Askin, 2000). In the oldest core, CRP-3, early Oligocene floras are of generally low diversity and are dominated by Nothofagidites, with common podocarpaceous conifers, a low diversity of other angiosperms, and rare cryptogams. The main elements of the floras are similar to those from the Prydz Bay Eocene, but there are differences in the minor components (e.g., in the presence of ?Stylidiaceae and Casuarinaceae). Raine and Askin (2000) suggested comparisons between the parent vegetation and low-diversity Nothofagus forests growing now in the magellanic region of southern South America, where low forests give way to Nothofagus scrubland above the altitudinal treeline.
Unlike East Antarctica, microfloras recovered from outcrops on the Antarctic Peninsula are likely to be in situ and incorporate relatively few recycled taxa. As such these provide the nearest equivalent of a continuous record for Cretaceous-Eocene time in Antarctica. Three areas that preserve diverse elements of the Transantarctic dinocyst floras are Cockburn Island (Hall, 1977), James Ross Island (Dettmann and Thompson, 1987), and Seymour Island (Wrenn and Hart, 1988; Askin, 1990a, 1990b; Askin et al., 1991). Many of the Eocene assemblages include recycled Late Cretaceous and Paleocene dinoflagellates not recorded at Prydz Bay.
The youngest diverse dinocyst floras come from the La Meseta Formation of Seymour Island. These late early to late middle-late Eocene assemblages (fig. 10 in Wrenn and Hart, 1988) provide further information regarding circum-Antarctic endemism and the age of the Prydz Bay assemblages in several ways: (1) a number of characteristic species, including A. antarcticum, D. antarctica, and E. partridgei are restricted to the late early Eocene interval; (2) none of the illustrated Deflandrea specimens appear to be D. "prydzensis;" and (3) Lejeunocysta and Octodinium spp. are restricted to the late middle-late Eocene interval.
Terrestrial palynofloras from Cockburn Island closely resemble the Prydz Bay microfloras in being dominated by gymnosperms (chiefly P. mawsonii, Podocarpidites spp., and Nothofagidites spp.). Other angiosperms make up ~5% of the microflora, but individual taxa, including Proteacidites, are uncommon to rare, as are cryptogams.
For the Cretaceous, diverse microfloras of that age are preserved on Dundee, James Ross, Seymour, and Vega Islands near the tip of the Antarctic Peninsula and on Seymour Island (Askin, 1988, 1990a, 1990b; Dettmann and Thompson, 1987; Askin et al., 1991). These range in age from Barremian to Maastrichtian and include many of the dinocysts used as zonal indices in southern Australia by Helby et al. (1987).
Comparison with the Cretaceous sequence at Site 1166 is hindered by the apparent absence of fossiliferous Turonian-Santonian material (see fig. 2 in Dettmann and Thompson, 1987) and the fact that most of the taxa that are shared are long-ranging spore species. Conversely, age-diagnostic dinocysts and pollen taxa found on the Antarctic Peninsula are absent from Prydz Bay or are present only as recycled elements. It may be that correlative sediments (i.e., of late Campanian age) do not occur in subcrop in the Prydz Bay catchment.