We selected our raw samples during the cruise by carefully examining each turbidite sequence indicated by sedimentary structures. Most samples were collected from the hemipelagic and pelagic layers situated at the top of the turbidite sequences in order to yield abundant and well-preserved Pliocene-Pleistocene nannofossil assemblages and to avoid severe reworking as far as possible. Smear slides were prepared directly from raw samples and were examined using phase contrast and polarized light microscopy to define the relative abundance of each nannofossil species present. A scanning electronic microscope (SEM) was employed to identify some biostratigraphically important species, such as Emiliania huxleyi and Reticulofenestra asanoi.
The relative abundance of individual species and the total abundance for each sample were tabulated on the distribution charts (Table 2, Table 3, Table 4, Table 5, Table 6, Table 7, Table 8, Table 9) using a light microscope with a magnification of ×1560. The letters used on the range charts and the corresponding definitions are as follows:
R = rare (1 specimen per 51-200 fields of view);
F = few (1 specimen per 11-50 fields of view);
C = common (1 specimen per 2-10 fields of view);
A = abundant (1-10 specimens per field of view); and
V = very abundant (>10 specimens per field of view).
Lowercase letters are used to denote reworking.
The preservation of calcareous nannofossils was recorded as follows:
G = good; individual specimens exhibit little or no dissolution or overgrowth; diagnostic characteristics are preserved and nearly all of the specimens can be identified;
M = moderate; individual specimens show evidence of dissolution or overgrowth; some specimens cannot be identified to the species level; and
P = poor; individual specimens exhibit considerable dissolution or overgrowth.
Calcareous nannofossil species considered in this paper are listed in the Appendix, where they are arranged alphabetically by generic epithets. Bibliographic references for these taxa can be found in Perch-Nielsen (1985) and Sato and Takayama (1992).