Pleistocene Collapse of the West Antarctic Ice Sheet

Alley R. B., Whillians I. M., Science 254, 959 (1991);

MacAyeal D. R., Nature 359, 29 (1992);

Bentley C. R., Science 275, 1077 (1997);

Bindschadler R., ibid. 276, 5313 (1997);

; R. A. Warrick et al., in Second Assessment Report of the Intergovernmental Panel on Climate Change: Working Group I, J. T. Houghton et al., Eds. (Cambridge Univ. Press, Cambridge, 1996).

Mercer J. H., Nature 271, 321 (1978);

. The exact elevation of sea level during peak MIS 5e is still somewhat unclear. Recent well-dated records indicate a complex sea level history for MIS 5e, characterized by typical elevations of 2 m and at least one relatively brief eustatic high stand of at least 6 m [

Neuman A. C., Hearty P. J., Geology 24, 775 (1996);

Kindler P., Strasser A., ibid. 25, 1147 (1997);

; C. E. Sherman et al., ibid.21, 1079 (1993)].

Scherer R. P., Palaeogeogr. Palaeoclimatol. Palaeoecol. 90, 359 (1991);

; J. Glaciol. 39, 716 (1993); L. H. Burckle, ibid., p. 491.

D. M. Harwood, R. P. Scherer, P.-N. Webb, Mar. Micropaleontol. 15, 91 (1989); R. P. Scherer, thesis, Ohio State University, Columbus, OH (1992).

A. Eisenhauer et al., Earth Planet. Sci. Let. 124, 171 (1994);

You C. F., et al., Geochim. Cosmochim. Acta 58, 4887 (1994);

Speilhagen R. F., et al., Geology 25, 783 (1997);

; A. Aldahan, S. Ning, G. Possnert, J. Backman, K. Boström, Mar. Geol. 144, 147 (1997).

A. Aldahan and G. Possnert, Quat. Sci. Rev., in press.

Alley R. B., Blankenship D. D., Bentley C. R., Rooney S. T., Nature 322, 57 (1986);

; D. D. Blankenship, C. R. Bentley, S. T. Rooney, R. B. Alley, ibid., p. 54; H. Engelhardt,

Humphrey N., Kamb B., Fahnstock M., Science 248, 57 (1990).

S. Tulaczyk, B. Kamb, R. P. Scherer, H. Engelhardt, J. Sediment. Res. 68, 487 (1998).

We estimated the absolute abundance of diatoms using a modification of a recognized method [R. P. Scherer, J. Paleolimnol. 12, 171 (1994)]. Abundance is normally estimated on the basis of whole diatoms, with the use of standardized counting criteria, but whole diatoms are too rare in these sediments to allow statistically significant estimates. Instead, we counted diatom fragments in the size class from 2 to 5 μm (Table 1). The abundance of whole diatom valves is at least three orders of magnitude lower. Extracting identifiable diatoms from these diamictons for biostratigraphic application requires specialized laboratory methods and extreme care to avoid sample contamination. All glassware and reusable supplies used in sample preparation are soaked in warm 15% NaOH for 4 hours and repeatedly rinsed with distilled and filtered water to remove any diatom residues that may be present. To test the reproducibility of abundance estimates and to test for potential laboratory contamination, we routinely prepare and analyze both sample duplicates and blanks. Duplicate preparations demonstrated consistent and reproducible results in diatom abundance (<10% error) and assemblage analyses. Thorough analysis of blanks revealed no evidence of sample contamination by identifiable diatom fragments.

AMS analysis of ¹⁰Be was performed at the Uppsala Tandem Laboratory with a machine and statistical error of <5% (1 SD); the error may become >5% because of boron correction and correction for the NIST SRM 4325 standard. ¹⁰Be was extracted from 0.5 to 1 g of a dried silt-clay fraction, using 200 μg of Be carrier, total dissolution (with hydrochloric acid and hydrofluoric acid), and cation column separation. ¹⁰Be (measured in atoms per gram) was normalized for the total sample dried at 110^oC for 6 hours. The silt-clay fraction used is dominated by quartz, with lesser quantities of feldspar and clay minerals and negligible carbonate (8).

H. Engelhardt and B. Kamb, J. Glaciol.43, 207 (1997).

Ice samples [M. Jackson and B. Kamb, J. Glaciol. 43, 415 (1998)] were washed under a stream of distilled deionized water to remove approximately 0.5 cm of the outer layers, then melted in new nylon beakers and filtered onto 25-mm microfilters. Additionally, two water blanks were filtered and analyzed (9). Filters were mounted on cover glass with Norland optical adhesive and analyzed in their entirety at ×1200 magnification. The age of the ice samples is unknown but is likely to be early Holocene.

This sample was taken during field season 1988–1989 (borehole 2, sample 1, Winnower sample). Much of the clay and other fine-grained material was lost in recovery of this sample (3).

This percentage is an underestimate because it does not include the many extant antarctic diatoms that were also common in the Miocene ocean.

Although it is ubiquitous in modern antarctic continental shelf sediments, the age of the first stratigraphic occurrence of T. antarctica is not yet well documented because of a lack of continuous Quaternary near-shore reference sections. Its appearance can be reliably bracketed by its occurrence in MIS 5e interglacial deposits [B. L. Ward and P.-N. Webb, J. Foram. Res. 16, 176 (1986)] at Cape Barne, Ross Island [R. P. Scherer, unpublished data] and by its absence from well-dated (0.75 My) lower Quaternary diatomaceous deposits in the Southern Ocean [

Gersonde R., Barcena M., Micropaleontology 44, 84 (1998);

] and in McMurdo Sound (Fig. 1) (Cape Roberts Project core 1, unit 3.1, upper A. ingens zone) [Cape Roberts Science Team, Terra Antarctica 5, 1 (1998)]. J. Barron [U.S. Geol. Surv. Open File Report 96-173 (1996)] reports T. cf. antarctica in mid-Pliocene Antarctic sediments, but we believe that there are significant taxonomic differences between T. cf. antarctica sensu Barron and T. antarctica sensu stricto, which occurs in our samples.

G. Brancolini, A. K. Cooper, F. Coren, in Geology and Seismic Stratigraphy of the Antarctic Margin, A. K. Cooper, P. F. Barker, G. Brancolini, Eds., vol. 68 of Antarctic Research Series (American Geophysical Union, Washington, DC, 1995), p. 209; D. E. Hayes and L. A. Frakes, Init. Repts. Deep Sea Drill. Proj. 28, 912 (1975).

The distance is reported with respect to the bed, compensated for ice stream flow between 1988 and 1995.

Rooney S. T., Blankenship D. D., Alley R. B., Bentley C. R., J. Geophys. Res. 92, 8913 (1987).

The Chaetoceros spore occurs in the sample containing the highest concentration of clothing fibers, thus the highest potential for sample contamination.

Kellogg D. E., Kellogg T. B., Geology 24, 115 (1996).

A massive volume of Pliocene and Pleistocene sediment has been eroded and transported toward the continental shelf edge by glacial processes [A. K. Cooper et al., Mar. Geol.102, 175 (1991);

Scherer R. P.Antarct. J. U.S.29199411] (16).

Howard W. R., Nature 388, 418 (1997);

; D. A. Hodell, Paleoceanogr. 8, 47 (1993).

MacAyeal D. R., Nature 359, 29 (1992).

Bindschadler R., Vornberger P., Science 279, 689 (1998);

Doake C. S. M., Corr H. F. J., Rott H., Skvarca P., Young N. W., Nature 391, 778 (1998);

Rott H., Skvarca P., Nagler T., Science 271, 788 (1996).

Support for this study came from Uppsala University, the Swedish Natural Sciences Research Council, and the Knut and Alice Wallenberg Foundation to R.P.S., A.A., and G.P., and from a U.S. NSF grant OPP9319018 to B.K. and H.E.