RESULTS

Our record of alkenone concentration and SST at Site 1020 shows dominant 100-k.y. glacial-interglacial cyclicity over the last 780 k.y. Global ice volume has a strong 100-k.y. component, and comparisons of our data with benthic ¹⁸O should enable us to examine phase relationships between alkenone-based SST, alkenone concentration, and ¹⁸O. Over the last four major glacial-interglacial cycles, glacial temperatures have reached remarkably consistent minima (~6°C). These temperatures exist today in the Gulf of Alaska (Levitus, 1994). Deglacial warming has averaged 7°-10°C over this period; however, peak interglacial temperatures show larger variation than full glacial conditions. Alkenone temperature estimates suggest that three of the last four interglacial periods have been 2°-3°C warmer than the Holocene. Earlier in the record, the average index increases noticeably in a manner consistent with overall warmer marine temperatures in the first half of the Brunhes Chron. Peak interglacial temperatures do not appear any warmer than the interglacials of the latest Pleistocene, and glacial sea-surface cooling was reduced (Fig. 5). Despite the overall similarity between alkenone SST and global ice-volume records, the timing of coldest SSTs along the northern California margin precedes greatest global ice volume by ~8-10 k.y. For example, during the last glacial cycle at Site 1020, minimum SSTs occurred at ~30 ka (OIS 3), ~9 k.y. before the LGM. A similar, though less well-constrained, early warming occurred at Site 1020 in each of the last four glacial-interglacial cycles.

Alkenones derive exclusively from phytoplankton, hence their concentration in marine sediments reflects the supply and preservation of marine organic matter. Studies have shown that organic carbon supply plays a more important role than preservation in determining the total fraction of marine C_org in sediments (Karlin et al., 1992; Reimers et al., 1992; Dean et al., 1994). Total organic carbon measurements are useful in assessing marine productivity rates (Müller and Suess, 1979; Eppley and Peterson, 1979; Rabouille and Gaillard, 1991); however, terrigenous C_org may comprise a significant portion of the total C_org in continental margin sediments. To develop a history of paleoproductivity along the northern California margin, we must estimate the marine fraction of C_org over time. We do not expect SC₃₇ and marine C_org to correlate perfectly because other marine plankton groups may contribute significant fractions of the total C_org. A comparison of SC₃₇ and SST shows that higher alkenone concentrations occur during interglacials than during glacial periods (Fig. 6). Although the time series of SC₃₇ and CaCO₃ (Lyle et al., Chap. 11, this volume) are not exactly in phase, comparison of the two records does show some significant similarities (i.e., shared peaks at ~68, ~260, ~360, and ~500 ka). We explore later the relationship between lower inferred coccolithophorid productivity and cold glacial SST estimated at Site 1020. On longer time scales, a general shift to higher SC₃₇ values occurs before OIS 13 (~500 ka using our preliminary time scale).