Major reorganizations of North Atlantic circulation have occurred in the past on the millennial time scale. These oceanographic changes, termed Dansgaard-Oeschger (D-O) cycles, are particularly evident in the sedimentary record as color changes in sediment deposited rapidly during oxygen isotope Stages 3 through 5 (Dansgaard et al., 1993; Bond et al., 1993). The color change is partly linked to variations in carbonate and hematite content (e.g., Barranco et al. 1989; Keigwin and Jones 1994; Adkins et al. 1997).
The Bermuda Rise is located in the northwest Atlantic Ocean approximately 1000 km from the continental margin of the eastern United States (Fig. F1). The rise is a depocenter for sediment redistributed from elsewhere (e.g., Biscaye and Eittreim, 1977; Laine and Hollister, 1981). Such sediment focusing has resulted in linear sedimentation rates up to 100 times the open-ocean average, allowing high-frequency changes in sediment composition to be preserved. Previous workers (Keigwin and Jones, 1989; Adkins et al., 1997) have shown that sub-Milankovitch variations in grayscale and CaCO3 content related to D-O cycles are evident in Bermuda Rise sediments.
Site 1063 was drilled on the Bermuda Rise during Leg 172 (Fig. F1). The upper 35 m of sediment at this location shows high-frequency cycles in P-wave velocity and bulk density apparently related to D-O cycles (Fig. F2A) (Keigwin, Rio, Acton, et al., 1998). However, an intriguing observation is that P-wave velocity is at times in phase and at times out of phase with bulk density (Fig. F2A). These physical properties changes are not consistent with simple variations in carbonate content as the physical cause for D-O cycles in sediment. The purpose of this paper is to determine the physical cause of D-O cycles in sediment at Site 1063.