CONCLUSIONS

We present here the 780-k.y. -SST and alkenone productivity record at Site 1020. We have made the following conclusions:

1. Glacial-interglacial SST amplitudes have varied consistently over the last 450 k.y.; however, the total amplitude dampens during the early Brunhes Chron. Also, interglacial SSTs are generally warmer than Holocene values by ~3°C.
2. Our Holocene-LGM SST decrease of 5°C is within the error of previous multiproxy studies such as Prahl et al. (1995), Sabin and Pisias (1996), and Ortiz et al. (1997).
3. The sea-surface glacial cooling of 5°C at Site 1020 may be the result of the complex interaction of this current system with sea ice, glacial atmospheric effects, and a southward expansion of the Alaskan Gyre in the glacial northeast Pacific. Our Holocene -based SSTs most closely approximate modern winter SST (or subsurface production); therefore, the only possible bias (relative to today) would involve a shift to a warmer season of production.
4. We infer a tight correlation between warmest alkenone SST and minimum ice volume over the last 800 k.y. This cannot be tested adequately until a suite of ¹⁸O data from Site 1020 at the same resolution become available. We observe similar relationships between alkenone's SST and ¹⁸O data in cores to the south (Herbert et al., 1995; T.D. Herbert et al., unpubl. data).
5. The paleoceanographic record at Site 1020 shows evidence for enhanced Haptophyte productivity during periods of warm SST and diminished productivity during cold SST. This inferred productivity pattern is consistent with sedimentary C_org content and with climatic information from redwood pollen (Heusser, Chap. 20, this volume).