The dinoflagellate cysts in these samples appear to be autochthonous, with minimal evidence of reworking from older material. The cysts were recovered mostly within the Miocene part of the sequence and were dominated by congruentidiacean, now protoperidiniacean, cysts. These are not well known in the published dinoflagellate literature; it is suspected that some may be new to science, and they are being studied further to provide detailed taxonomic descriptions. Virtually all the dinoflagellate cysts recovered can be assigned to or associated with the modern heterotrophic dinoflagellate Protoperidinium. By comparison with our knowledge of the modern cyst flora, which also is dominated by Protoperidinium cysts (Harland et al., 1998), a similar cold-water setting with high nutrient content can be interpreted for the sampled horizons. It is most probable that these dinoflagellates were feeding upon diatom populations that were taking advantage of seasonality and high supply of nutrients. The lack of autotrophic gonyaulacacean dinoflagellate cysts is noteworthy.
Similar situations, especially within the Miocene, have also been recognized elsewhere in the world, for example, at high latitudes in both the Labrador Sea and Baffin Bay (Head et al., 1989a, 1989b, 1989c), in the Gulf of Mexico (Duffield and Stein, 1986), and from the eastern seaboard of the United States (de Verteuil and Norris, 1992).
The relationship between diatom content and dinoflagellate cyst content of the samples, however, is not straightforward (Fig. F2). Many diatomaceous samples contain few or no dinoflagellate cysts. Diatoms were most abundant from 4 to 5.3 Ma (see Pudsey, Chap 25, this volume), whereas dinoflagellate cysts attained highest abundance and diversity between 7 and 8 Ma. Dinoflagellate cysts persist at the base of the section where diatoms become less common because of silica dissolution. In the pairs of samples from bioturbated and laminated facies, the bioturbated sample tends to be the poorer in dinoflagellate cysts (e.g., samples at 4.20 Ma compared with 4.14 Ma, 5.80/5.85 Ma, 6.25/6.26 Ma, and 8.65/8.64 Ma). Foraminiferal linings are also less common in the bioturbated samples so this may simply be a preservational bias.
Finally, all the samples are almost barren of terrestrial palynomorphs such as angiosperm pollen, bisaccate grains, and spores, and only two of the samples contained any reworked material.