Significant quantities of fine-grained aragonite are produced on the shallow tops of the Bahama Banks when they are flooded during Quaternary highstands of sea level (Boardman and Neumann, 1984; Droxler et al., 1983). The aragonite is suspended and transported from the bank tops by tides and storms and settles to the seafloor, together with the tests of planktonic organisms. This produces expanded highstand sediment sections with very rapid sedimentation rates (up to 100's of cm/ka) within a normal marine-oxygen-isotopic record.
As aragonite produced in shallow-water contains up to 6ppm U and as the highstand sediment consists of up to 90% aragonite, these sediments contain orders of magnitude more primary U than other deep-sea sediments and therefore offer considerable potential for direct U/Th dating of the marine-oxygen-isotope record. In addition, because the water depths are shallow and sedimentation rates are high, very little Th is scavenged in the water column (so 230Th ingrown from U is not masked by scavenged Th). The rapid sedimentation rate also dilutes the clay content of the sediment so detrital U and Th concentrations are small, and means that sediment mixing is not expected to be significant.
To test the potential of this material for U/Th dating and to directly date the timing and duration of oxygen-isotope substage 5e we have performed U and Th isotope analyses on the mud fraction of sediments from the slope of Little Bahama Bank. Two shallow-water (×500m) cores were selected from topographic highs on the basis of their carbonate mineralogy and d18O stratigraphy. d18O analyses were performed on the bulk mud fraction and on coexisting G. sacculifer and the close agreement between the two stratigraphies indicates that the mud fraction and foraminifera are of the same age. A suite of stage-5 samples, together with two samples each from the late Holocene and stage-7, were selected for U and Th analysis by TIMS. Corrections to the U-series ages for detrital U and Th are made using sediment Al concentrations measured by neutron activation. Corrections for scavenged Th are made using the measured 232Th/230Th ratios. Both of these corrections are small, however, so we have a high degree of confidence in the validity of the corrected ages. This confidence is increased by the fact that the corrected Holocene ages are in broad agreement with 14C ages on the same samples and that substage-7a ages are in good agreement with previous estimates of the age of this stage.
Eight U/Th ages for stage-5e sediment range from 120.7-126.9 ka with uncertainties ranging from 2.7 - 6.5 ka (propagated to include uncertainty due to the corrections as well as analytical reproducibility). These dates, the first direct U/Th dates on submarine sediments, are thus in close agreement with the Specmap timescale (Imbrie et al., 1984). No evidence is seen for peak interglacial conditions preceding the insolation maximum and no evidence is seen for a long duration (>12ka) interglacial maximum as seen in some other records (e.g.Devil's Hole (Winograd et al., 1992)) . These ages therefore support the commonly accepted Milankovitch timescale for the marine oxygen-isotope record and demonstrate the potential of using banktop-derived aragonite sediment for further dating of the marine sediment and d18O records.
Boardman, M. R. & Neumann, A. C., Jour. Sed. Petrol. 54, 1110-1123 (1984).
Droxler, A. W., Schlager, W. & Wallon, C. C., Geology 11, 235-239 (1983).
Imbrie, J. et al., In Milankovitch and Climate (Berger, A. et al., eds.) 269-305 (Reidel, Dordrecht, Netherlands, 1984).
Winograd, I. J., Coplen, T. B., Landwehr, J. M., Riggs, A. C., Ludwig, K. R., Szabo, B. J., Kolesar, P. T. & Revesz, K. M., Science 258, 255-260 (1992).