The establishment of a Cenozoic biostratigraphy succession of the western Barents Shelf has been hampered by extensive reworking, especially in the younger sequences. Several controversies over the ages of the Cenozoic units have been caused by misinterpretation and miscorrelation of reworked microfaunas and floras, which frequently comprise the dominant part of the assemblages. In the examined material from Site 986, the high number of species recorded (see range charts in Appendix A and Appendix B) is a result of the dominant reworking of older Cenozoic and Mesozoic taxa. This is true for the whole cored succession and is especially typical for the units above Reflector R7. The relative abundances of reworked vs. in situ dinoflagellate cysts are shown in Figure 4 and Figure 5.
The most commonly reworked dinoflagellate cysts in the upper Pliocene-lower Pleistocene of Hole 986D (Fig. 5) are of Cretaceous age. One interesting feature is the presence of Upper Cretaceous taxa at several levels (including Diconodinium arcticum, Elytrocysta druggii, Heterosphaeridium difficile, Trithyrodinium verrucosum, Chatangiella granulifera, Chatangiella ditissima, Chlamydophorella? grossa, and Isabelidinium bakeri). Today, Upper Cretaceous strata are missing on Svalbard, but reworked Upper Cretaceous dinoflagellate cysts are relatively common in the Paleogene sequences on Sørkapp Land on Spitsbergen. Reworked Paleogene forms are generally less frequent than Cretaceous taxa, but a peak abundance comprising 25% of the total dinoflagellate assemblage occurs in the sample at 679.94 mbsf. Jurassic dinoflagellates are rarer and in most samples comprise <2% of the total assemblages. A distinct acme of recycled Jurassic dinoflagellate cysts is, however, observed at 718.34 mbsf (including the species Chlamydophorella ectotabulata, Chytroeisphaeridia cerastes, Paragonyaulacysta calloviense, Perisseiasphaeridium pannosum, and Sirmiodinium grossii). The reworked Jurassic cysts are dominantly of post-Bathonian age and are probably derived from deposits of the Fuglen and Hekking Formations from the western Barents Shelf and/or the Janusfjellet Formation on eastern Svalbard.
The distribution pattern of reworked dinoflagellate cysts in the Pleistocene sediments of Hole 986C (Fig. 4) is broadly comparable to that of 986D. Recycled Cretaceous taxa are common, including both Lower and Upper Cretaceous species. Reworked Paleogene forms are most common above 200 mbsf but show a highly fluctuating distribution pattern. Jurassic specimens are generally less frequent than reworked Cretaceous and Paleogene forms but more common than in the older strata of Hole 986D. Distinct peak abundances of recycled Jurassic dinoflagellate cysts are recorded at 276.94 (28%), 267.34 (26%), 132.03 (16%), and 122.18 mbsf (19%). In contrast to Hole 986D, reworked Miocene dinoflagellates are found above 267.34 mbsf in Hole 986C, where they occur together with Paleogene and Mesozoic cysts. They are, however, rare and comprise only 1% to 2% of the total assemblage when present. A small increase in the relative abundance of recycled Miocene dinoflagellates (4%) is found in the uppermost studied sample at 3.74 mbsf.
In their study of the Miocene-Pleistocene of the De Soto Canyon, in the Gulf of Mexico, Wrenn and Kokinos (1986) recognized a correlation between reworked dinoflagellate cysts, carbonate fluctuations, and glacial cycles. The occurrence of reworked pre-Miocene dinoflagellate cysts was directly related to the abundance of clay/silt succession and inversely related to the carbonate content and coarse fraction percent. They concluded that if glacial-interglacial cycles determined the relative proportions of silt, clay, and carbonate in the Pliocene-Pleistocene of the Gulf of Mexico, then these cycles also controlled the influx of reworked dinoflagellate cysts. Because of the low sampling frequency, it is not possible to determine in detail the relationship between lithology and the relative proportions of reworked dinoflagellate cysts at Site 986. Obviously, however, there is a cyclicity in the abundance of both in situ and reworked dinoflagellate cysts in the upper Pliocene-Pleistocene (Fig. 4, Fig. 5). More data are needed before this can be evaluated further.
One interesting aspect is that the preservation of the reworked Mesozoic dinoflagellate cysts is generally very good. In contrast, dinoflagellate cysts from the Jurassic and Cretaceous formations on western and southern Spitsbergen are usually poorly preserved and have undergone moderate to high thermal maturation. Nonetheless, dinoflagellates from the Jurassic and Cretaceous of Kong Karls Land to the east in the Svalbard region, on Franz Josef Land, and on the shallow eastern parts of the Norwegian Barents Shelf are generally well preserved. This suggests a transport direction from the east toward the shelf margin and then following a northwest direction similar to the present West Spitsbergen Current (Fig. 6). Dinoflagellate cysts behave like fine silt particles in the water column and may be transported considerable distances in the ocean. A provenance area on the central to eastern Barents Shelf or eastern Svalbard, following a westerly drainage route toward the major depositional fans along the western shelf margin, is thus likely.
Willard (1996) suggested that reworked palynomorph assemblages from the Pliocene-Pleistocene at Sites 910 and 911 on the Yermak Plateau had their source on the Eurasian shelf, where they were either entrained within sea ice and transported via the Transpolar Drift or were conveyed in surface currents following a similar path. This assumption was based on the relationship of the reworked Upper Cretaceous pollen and spore assemblages to contemporaneous floras of the boundary between the Normapolles and Aquilapollenites Provinces in the Laptev Sea area. Upper Cretaceous terrestrial palynofloras comparable to those found reworked at Sites 910 and 911 are, however, also found in the marine Upper Cretaceous formations of the Norwegian Barents Shelf (i.e., Maud and Tromsø Basins). Reworked Aquilapollenites are also observed in the upper Pliocene-Pleistocene deposits at Site 986. The present observations suggest that the upper Cretaceous palynofloras and most of the other reworked Mesozoic palynomorphs at Site 986 came from the Barents Shelf via the main drainage areas into the Bjørnøya and Storfjorden Troughs (Nøttvedt et al., 1988) and were then transported with the West Spitsbergen Current into the depositional site west off Spitsbergen (Fig. 6).