(Table 1) Sea-surface temperatures, intermediate-water temperatures, and latitudinal gradient at three time slices at the end of the Cretaceous

Climatic and oceanographic variations during the last 2 m.y. of the Maastrichtian inferred from high-resolution (10 k.y.) stable isotope analysis of the mid-latitude South Atlantic Deep Sea Drilling Project Site 525 reveal a major warm pulse followed by rapid cooling prior to the Cretaceous-Tertiary boundary. Between 66.85 and 65.52 Ma, cool but fluctuating temperatures average 9.9 and 15.4°C in intermediate and surface waters, respectively. This interval is followed by an abrupt short-term warming between 65.45 and 65.11 Ma, which increased temperatures by 2-3°C in intermediate waters, and decreased the vertical thermal gradient to an average of 2.7°C. This warm pulse may be linked to increased atmospheric pCO2, increased poleward heat transport, and the switch of an intermediate water source from high to low-middle latitudes. During the last 100 k.y. of the Maastrichtian, intermediate and surface temperatures decreased by an average of 2.1 and 1.4°C, respectively, compared to the maximum temperature between 65.32 and 65.24 Ma.

Stable carbon isotope ratios of Cibicidoides wuellerstorfi of MIS 2 time slices

The interval of time represented by marine isotope stages 11 and 12 (~360-470 ka) contains what may be the most extreme glacial and interglacial climate conditions of the Late Pleistocene. It has been suggested that sea level rose by ~160 m at the termination of glacial stage 12. This is 30% greater than the sea level rise that followed the most recent glacial maximum. There have been few detailed studies of the unique conditions that existed during the stage 11-12 time period because of the lack of high-quality core material. This problem has been addressed by the collection of high deposition rate cores from sediment drifts in the western North Atlantic during Ocean Drilling Project Leg 172. Benthic foraminiferal d13C data from cores collected between ~4600 and 1800 m were used to reconstruct bathymetric gradients in deep and intermediate water properties for selected time slices during this glacial-interglacial cycle. During glacial stage 12, the deep western North Atlantic was filled by a water mass that was more nutrient-enriched than modern Antarctic Bottom Water. Above 2000 m, a more nutrient-depleted water mass existed during this glacial stage. Such an intermediate water mass has been described for more recent glacial periods and presumably forms in a more proximate region of the North Atlantic. Interglacial stage 11 water mass properties closely resemble those of the present-day western North Atlantic. A nutrient-depleted water mass (d13C of 0.75-1.0 per mil), similar to modern North Atlantic Deep Water existed between 3500 and 2000 m. This was underlain by a water mass with lower d13C values (<0.75 per mil) that probably was derived from a southern source. Using Leg 172 data, along with previously published results from the Atlantic and Pacific oceans, we estimate a mean global d13C change of 0.95 per mil from stage 12 to stage 11. This is twice the whole ocean ?13C change reported for the transition from the last glacial maximum to the Holocene.

Sediment Trap Data from the CARIACO Ocean Time Series (1995-2010)

The Cariaco Basin is a 1400-m-deep depression approximately 160 km long by 70 km wide located off the central Venezuelan coast . It is connected to the Atlantic Ocean by a sill ~100-m-deep, and two slightly deeper channels that breech it; Canal Centinela (146-m-deep) and Canal de la Tortuge (135-m-deep). High surface production rates and restricted circulation result in anoxic waters below ca. 275 m. The depth of the oxycline varies between 250 and 320 m and is independent of density. Rather, fluctuations in oxycline depth appear to be due to lateral intrusions of Caribbean Sea water that are linked to eddies along the continental shelf. A mooring with five sediment traps (Z, A-D) is located in the eastern Cariaco Basin. Traps A-D have been in place since November 1995. Trap A is located in oxic waters at 226 ± 6 m. Trap B is located at 407 ± 3 m and Trap D is located at 1205 ± 3 m. Trap C was located at a depth of 880 ± 2 m from Jan. 1996 to Nov. 2000, and was moved to 807 ± 2 m in Nov. 2000. A fifth trap, Z, was added in November 2003 at 110 m for the first 6 months, and at 150 m thereafter. All five sediment traps are coneshaped with a 0.5 m**2 opening that is covered with a baffle top to reduce turbulence. The mooring is deployed for six-month intervals and each sample collection cup is filled with a buffered 3.2% formalin solution as a preservative for the accumulating organic matter. The cups are numbered 1-13, with cup 1 collecting for the two-week interval immediately following deployment, and cup 13 collecting for the 2 weeks immediately before recovery.