Before the Pleistocene, climatic rhythms producing Northern Hemisphere glaciations were dominated by a 41-k.y. frequency (Fig. 4A; Ruddiman et al., 1986, Raymo et al., 1990). Whole-core magnetic susceptibility data, however, reveal that in the Indian Ocean, Pliocene climate variability before 2.5 Ma responded to precessional changes in insolation with periodicities of 23 and 19 k.y. (Bloemendal and deMenocal, 1989). Sediments and isotope data from the low-latitude Bahamas (Leg 166) indicate a similar discrepancy between the sedimentary and the isotope records. Oxygen isotopes record sea-level changes at an obliquity frequency, whereas resistivity and gamma-ray values record sea-level changes dominated by orbital precession (Fig. 4).
Spectral analysis of resistivity and gamma-ray data of marl/limestone alternations from the Santaren Channel (Leg 166) show that the sedimentary rhythm of marl/limestone alternations is in concert with orbital precession, indicating sea-level changes of this frequency (Fig. 4). In contrast, the frequency spectrum of oxygen isotope data from planktonic foraminifers from the same core clearly displays a dominant obliquity frequency (Fig. 4A; Ruddiman and McIntyre, 1984; Raymo et al., 1990). We speculate that both frequencies of sea level existed throughout most of the Neogene.
However, there is also a strong peak in the power spectrum at 11 ka. For example, in the late Miocene section at Site 1003 (Fig. 4A), Kroon et al. (Chap. 15, this volume) report two resistivity log peaks within each precessional cycle in Unit 2 (3.6 to 4.6 Ma) at Site 1006, indicating an 11-k.y. duration of each cycle. Counting the middle Miocene cycles from 738 to 915 meters below seafloor at Site 1003 also rendered 11 k.y. duration (Bernet, 2000). Sedimentary cycles of 11 k.y. duration are known from the continental tropics and are attributed to climate changes within the precessional orbital cycle (Olson, 1990; Short et al., 1991). If the marl/limestone alternations were indeed the result of fluctuating sea level, it would imply that these climate changes produced high-frequency, low-amplitude (5-10 m) sea-level changes. The precessional signal is also present in the frequency spectrum from oxygen isotopes of Site 1006 (J. Wright, pers. comm., 1999) and as a minor peak in many oxygen isotope spectra (e.g., Raymo et al., 1990). The Earth's precession could theoretically lead to an alternating intensification of glacial buildup on the North and South Poles, causing small-scale sea-level changes with an 11-k.y. cyclicity. However, this cyclicity has not yet been detected in the oxygen isotope record of the Bahamas Transect.
Sea-level changes can be orbitally induced and created by changes in the ice volume at the poles. Oxygen isotopes indicate that obliquity forces created the largest changes in ice volume. However, smaller scale changes might occur and be recorded in shallow-water carbonates in the tropical realm. The shallow-water carbonates are sensitive indicators capable of recording sea-level changes of a few meters. Such changes require only a small amount of ice volume modification that might be undetected in oxygen isotope values of pelagic sediments because of low sedimentation rates. Where sedimentation rates are high, oxygen isotopes from the sediment drift in the Santaren Channel do indeed display precessional and obliquity periodicity in the frequency spectrum. Sea-level changes of precessional frequency probably transpire for most of the Neogene and dominate the sedimentary record in low latitudes. This result also explains the dominance of precession found in ancient shallow-water carbonate cycles. Site 1006, the most distal site of the transect, is positioned on a thick sediment drift (Fig. 1). These drift deposits provide a continuous and expanded section for most of the Neogene section. Fine biostratigraphy, the basis for a biocyclostratigraphy, has been established at this site (Eberli, Swart, Malone, et al., 1997; Wright and Kroon, Chap. 1, and Kroon et al., Chap. 15, both this volume). As mentioned above, the sedimentation in this drift on the slope of Great Bahama Bank is controlled by the precessional beat for most of the Neogene. Counting these cycles produces an accurate time scale and, thus, a sedimentation rate time scale (Kroon et al., Chap. 15, this volume). Spectral analysis of these sedimentation rates indicates a bundling of the precessional beats into longer term eccentricity cycles. The short-term eccentricity (~120 k.y.) and the long-term cycles of eccentricity (400 k.y.) are pervasive in the Miocene, but even the long-term 2-Ma eccentricity cycle is present (Kroon et al., Chap. 15, this volume). If these sedimentation rates are caused by changing sea level, this would be the first time that a mechanism could be documented by which the high-frequency cycles bundle into lower order (third-order) sea-level changes. Future work on stable isotopes will elucidate this process.