A total of 475 samples from 300.82 to 518.11 mcd at Site 1144 were used in this study. Sampling spacing varies from 30 cm between 300.82 and 415.58 mcd to 50–70 cm between 415.58 and 518.11 mcd. Three intervals lack samples because of either incomplete core recovery or contamination: 352–357 mcd, 365–370 mcd, and 387–391 mcd. On the basis of isotopic stratigraphy, the time resolution of these samples varies from 500–1000 yr.
Samples were processed with standard techniques. Residue >63 µm was collected and separated into two fractions using a 150-µm sieve. More than 300 planktonic foraminifers from a portion of the >150-µm fraction were picked and identified, and the percentage abundances of various species were calculated and charted. We followed the species taxonomy of Kennett and Srinivasan (1983), Bolli and Saunders (1985), and Hemleben et al. (1989). Species preferences to climate belts or faunal provinces, as summarized by Bé (1977), were used for grouping species.
Test fragments were also counted. Fragmentation results based on eight fragments equaling one complete test were used to infer carbonate dissolution in the bottom water (Le and Shackleton, 1992). Coarse fraction data show higher percentages in glacial intervals than in interglacial intervals, indicating stronger offshore transport at sea level lowstands (Fig. F2).
SST was estimated using paleoecology transfer function FP-12E of Thompson (1981) on planktonic foraminifer census data (see the "Appendix"). The method developed by Pflaumann and Jian (1999) was not employed because it often underestimates the winter SST because of few modern analogs from high latitudes. Figures F2, F3, F4, F5, and F6 illustrate the isotopic and coarse fraction results as well as the abundance profiles of common species documented in this study.