Description: The 1.4-m.y.-long stable oxygen isotope record of Site 1006 in the low-latitude North Atlantic Ocean shows large glacial/interglacial amplitude changes caused by a combination of temperature and salinity fluctuations. A trend of increased sea-surface temperatures during the interglacial periods is present in the record beginning at isotopic Stage 11 and ultimately leading to the lightest d18O values in isotopic Stages 9, 5, and 1. Maximum d18O values are recorded during glacial isotopic Stages 6 and 8. Stable isotopic variability increased during the Brunhes Chron at the 100-ka time scale. The large amplitude changes can best be explained by global and regional ocean circulation changes. Increased strengthened return flow of warm salty water from the Pacific may have occurred during interglacial periods since isotopic Stage 11, which was largely reduced during glacial periods. The large climate fluctuations had a profound effect on the shallow-water carbonate production of the Great Bahama Bank. The aragonite content of the sediments shows fluctuations that follow the d18O record. The leeward side of the Great Bahama Bank received increased input of platform material during sea-level highstands when the sea-surface waters were warm.

Description: We have performed U-Th isotope analyses on pure aragonite samples from the upper sections of Leg 166 cores to assign each aragonite-rich sediment package to the correct sea-level highstand. The uppermost sediment package from each of the four sites investigated (Sites 1003, 1005, 1006, and 1007) yielded a Holocene U-Th age. Sediment packages from deeper in the cores have suffered diagenesis. This diagenesis consists of significant U loss (up to 40%) in the site nearest the platform (Site 1005), slight U gain in sites further from the platform, and continuous loss of pure 234U caused by alpha recoil at all sites. The difference in diagenesis between the sites can be explained by the different fluid-flow histories they have experienced. Site 1005 is sufficiently close to the platform to have probably experienced a change in flow direction whenever the banks have flooded or become exposed. Other sites have probably experienced continuous flow into the sediment. Although diagenesis prevents assignment of accurate ages, it is sufficiently systematic that it can be corrected for and each aragonite-rich package assigned to a unique highstand interval. Site 1005 has sediment packages from highstands associated with marine isotope Stages 1, 5, 7, 9, and 11. Site 1006 is similar, except that the Stage 7 highstand is missing, at least in Hole 1006A. Site 1003 has sediment only from Stage 1 and 11 highstands within the U-Th age range. And Site 1007 has sediment only from the stage 1 highstand. This information will allow the construction of better age models for these sites. No high-aragonite sediments are seen for Stage 3 or Substages 5a and 5c. Unless rather unusual erosion has occurred, this indicates that the banks did not flood during these periods. If true, this would require the sea level for Substages 5a and 5c to have remained at least ~10 m lower than today.