The biostratigraphic zonation scheme used for the Pleistocene (Fig. F2) is that of Gartner (1977). The marker species Gartner uses are generally present and easily recognizable and did not require a detailed morphometric analysis of gephyrocapsids, which was beyond the scope of this study. Some markers were not consistently present, and additional events were used to approximate the Gartner (1977) zonal boundaries. The bottom of the Pseudoemiliania lacunosa Zone is marked by the end of the dominance of small Gephyrocapsa species by Gartner (1977). As this could not be accurately determined with qualitative methods and the alternate marker Gephyrocapsa parallela is not consistently present, the last occurrence (LO) of Reticulofenestra asanoi was used to approximate this boundary (Fig. F2). The LO of Helicosphaera sellii is used here to mark the top of the Helicosphaera sellii Zone, taking into account that this datum has been reported as diachronous (Backman and Shackleton, 1983), ecologically sensitive, and rare at the end of its distribution (Perch-Nielsen, 1985).
The Pliocene/Pleistocene boundary is defined by the LO of Discoaster brouweri (1.95 Ma) by Gartner (1977). This nannofossil definition of the Pliocene/Pleistocene boundary is 0.18 m.y. older than the magnetostratigraphically defined Pliocene/Pleistocene boundary at Termination C2n (1.77 Ma) (Berggren et al., 1995). Discoaster brouweri was recorded only at Site 1172; therefore, alternative nannofossil markers were used to approximate the Pliocene/Pleistocene boundary in the nannofossil biostratigraphy. At Site 1168, the first occurrence (FO) of Gephyrocapsa caribbeanica (1.72 Ma) was used to approximate the Pliocene/Pleistocene boundary. At Sites 1170 and 1171, the LO of Calcidiscus macintyrei (1.59 Ma) was used to roughly approximate the Pliocene/Pleistocene boundary, as G. caribbeanica and Gephyrocapsa oceanica (FO = 1.65 Ma) were discontinuously present.