Site 1018 (36°59´N, 123°17´W; water depth, 2477 m) is coastal and shallow, located on the sediment drift at the flank of the Guide Seamount, and the sedimentation rate is relatively high compared to those of the other two sites (157 m/m.y. on average).
Site 1020 (41°00´N, 126°26´W; water depth, 3038.4 m) is situated on the east flank of Gorda Basin and is ~170 km west from shore. This is a paleoceanographically well-defined site; the oxygen isotope record (A. Mix, pers. comm., 1998) and the calcium carbonate and biogenic opal records (Kuroda et al., Chap. 14, this volume) of the late Pleistocene have been already obtained.
Site 1021 (39°05´N, 127°47´W; water depth, 4211.5 m) is located on the Outer Delgada Fan, ~360 km from shore. The extremely low sedimentation rate (779 m/m.y.) and the siliciclastic clay sediments seem to reflect very low biogenic production in the sea surface above this drill site.
Samples for the C1n (Brunhes) Subchron were taken at roughly 70-cm intervals: time intervals for Sites 1018, 1020, and 1021 correspond to sampling intervals of ~5, ~8, and ~21 ka, respectively. Samples representing the early Pleistocene and the late Pliocene were taken at roughly 340-cm intervals down to Core 39X for Hole 1018A.
A set of strewn slides was prepared following the modified procedures used by Koizumi (1992) and Ikeda et al. (2000). After drying in an oven at 50°C for 24 hr, 0.1 g (±5%) of the sample was dissolved in 10 mL of hydrogen peroxide solution (15% strength) in a 100-mL beaker and boiled for several minutes. After filling with distilled water, the beaker was left to stand for 5 hr. Then the supernatant was poured off, and distilled water was poured in again. This decanting process was repeated four times to dilute acidity in the suspension. The residue was finally diluted with 100 mL of distilled water and homogenized for several seconds in an ultrasonic bath. Using a micropipette, 0.5 mL of this solution was spread on a square cover glass (18 × 18 mm) and dried on a hot plate at 50°C. The prepared cover glass was mounted on a glass slide using Pleurax.
All diatoms were identified and counted under a light microscope with magnification of 600× until the total number exceeded 200 (resting spores of Chaetoceros spp. not included). The counting was accomplished by observing up to six traverses spanning the entire width of the cover glass. The total number did not reach 200 specimens when diatoms were sparse. Because the weight of sediment on a cover glass is measurable, the abundance of diatom valves (number per gram) was obtained by using the numbers of transects and the total amount of diatom valves. The results are presented in Table 1, Table 2, Table 3, and Table 4.
Diatom abundance fluctuates significantly at Sites 1018 and 1020, whereas diatoms are very scarce in the section studied at Site 1021. In general, when the total number did not reach 200 specimens, representative diatom assemblages could not be obtained even by relative abundance (percentage of total diatom valves) because the value contains a statistically serious error. Therefore, all taxonomic abundances were indicated by absolute values (number of diatom valves per gram of sediment).
Judging conditions of preservation depended on the extent of dissolution, fragmentation, and overgrowth of diatom valves through observation with a light microscope.