Foreshadowing the glacial era

Article metrics

Under what circumstances do glaciations persist or occur only transiently? Indications that short-lived ‘icehouse’ conditions occurred during the otherwise warm Eocene provide further cause for debate on the question.

Earth entered its present glacial state 34 million years ago with the growth of the Antarctic ice sheet1. This major climate transition occurred abruptly and essentially irreversibly at the Eocene–Oligocene boundary, a conclusion based on the record of ice-sheet size preserved in the oxygen isotopic composition of limestones2,3. The preceding Eocene epoch (55–34 million years ago) is generally considered to have been warm and ice-free, but data on this time interval, as recorded in cores of marine sediments, have been sparse.

By analysing newly acquired core material from the tropical Pacific, Tripati et al.4 (page 341 of this issue) provide a much more detailed view of the climate system before the permanent transition to the glacial state. What they find there, and in contemporaneous cores from the South Atlantic, is several small glaciations and one major (but transient) glaciation in the middle to late Eocene, millions of years before the Eocene–Oligocene boundary.

The water from which continental ice sheets grow derives from evaporation of ocean water and its deposition at high latitudes as snow. Thus, as ice sheets grow, sea level falls. Moreover, compared with sea water, the snow is enriched in the lighter isotope of oxygen, 16O. So, as ice sheets grow, the ratio of 18O to 16O in the oceans increases; the ratio is generally presented as the standardized ratio δ18O. Organisms that precipitate skeletons of calcium carbonate (CaCO3) do so close to oxygen isotopic equilibrium with the waters in which they grow, so the δ18O of fossil skeletons provides a proxy measure for ice-sheet size in the past. However, as the equilibrium δ18O of the CaCO3 also depends on temperature, an unambiguous interpretation of ice-sheet size from fossil δ18O requires additional temperature information. Tripati et al.4 use an independent temperature proxy — the amount of magnesium incorporated into CaCO3 shells — to isolate the effects of changing ice volume.

The temperature proxy indicates that there was little global cooling associated with the late Eocene glaciations, suggesting that, as in the earliest Oligocene event5, most of the shift in δ18O was due to an increase in ice volume and that cooling may have been limited to high latitudes (or that ice-sheet accumulation there was limited by moisture rather than temperature). As in the earliest Oligocene, the isotopic data seem to require the presence of ice sheets on Antarctica at least as thick as those today, and substantial ice sheets in North America (most likely Greenland). This latter result runs contrary to conventional wisdom, which holds that the Northern Hemisphere glaciation began tens of millions of years later6.

From their analyses of accumulation patterns of CaCO3 on the sea floor, Tripati et al.4 find signs of substantial perturbations in the ocean's carbon cycle during the Eocene, patterns that mimic those of the more permanent change to come. The oceanic calcium-carbonate compensation depth (CCD), the depth below which CaCO3 does not accumulate because deep waters are corrosive, increased significantly during the glacial events. Deepening of the CCD is an expected consequence of sea-level fall, because it allows for additional deep-sea carbonate accumulation that compensates for the loss of carbonate deposition from shallow waters7. So these data corroborate the claim that substantial ice sheets existed in the Eocene.

Why the Oligocene Antarctic ice sheet persisted but the Eocene ice sheets did not is unclear. As a driver for glaciation, Tripati et al. invoke a reduction in the amount of atmospheric CO2 accompanying the growth of the Himalayas and resulting from enhanced chemical weathering of the rocks unearthed8,9. They suggest that increased biological productivity in the early Oligocene, and so increased use of CO2 in photosynthesis, may have provided the additional drawdown of atmospheric CO2 that was necessary to sustain the glaciation. Perhaps the stability of the glacial state increased as atmospheric CO2 levels fell, so that stochastic effects (such as volcanic eruptions releasing CO2, or destabilization of methane embedded in the sea floor), or variations in Earth's orbit, became insufficient to jar Earth out of its glacial state (Fig. 1). Future work on proxy measures of atmospheric CO2 from the Eocene and Oligocene should provide the necessary test for this hypothesis.

Figure 1: Glacial stability and instability.

Global temperature is indicated by the balls. The findings of Tripati et al.4 suggest that a glacial (‘icehouse’) climate state could have existed in the otherwise warm Eocene. But the ‘greenhouse’ state was perhaps readily re-established through instabilities arising from stochastic or orbital fluctuations (indicated by the arrow). As levels of atmospheric CO2 fell, the stability of the icehouse state might have increased and that of the greenhouse state might have decreased, so that by the early Oligocene any such fluctuations would not have moved the system back to the greenhouse state, but would have been followed by rapid coolings to the stable glacial state. (After Fig. 2–3 in ref. 10.)

The equatorial Pacific sediments analysed by Tripati et al.4 are thought to represent oceanographic conditions over a broad region of the Pacific, and the data from the South Atlantic support the proposition that these changes were indeed globally significant. Nevertheless, a general acceptance that glaciations occurred in the middle to late Eocene will probably require further evidence. The suggested existence of large Northern Hemisphere ice sheets in the Eocene is highly controversial. Moreover, the fidelity of the magnesium content of CaCO3 as a measure of temperature demands further scrutiny. However, the existence of precursor glaciations foreshadowing the major transition to the glacial state is theoretically expected of a system that is subject to natural fluctuations but is gradually evolving from one stable state to another.

If decreasing atmospheric CO2 stabilized the glacial state in the Oligocene, might increasing atmospheric CO2 from fossil-fuel burning destabilize it in the future? The lesson to be learned here is that we should watch for subtle signs that we are moving from the icehouse world in which Earth has remained for 34 million years into a new, greenhouse world.


  1. 1

    Wise, S. W., Breza, J. R., Harwood, D. M. & Wei, W. in Controversies in Modern Geology (eds Mueller, D., McKenzie, J. & Weissert, H.) 133–177 (Academic, San Diego, 1991).

  2. 2

    Kennett, J. & Shackleton, N. J. Nature 260, 513–515 (1976).

  3. 3

    Zachos, J. C., Quinn, T. M. & Salamy, K. A. Paleoceanography 11, 251–266 (1996).

  4. 4

    Tripati, A., Backman, J., Elderfield, H. & Ferretti, P. Nature 436, 341–346 (2005).

  5. 5

    Coxall, H. K. et al. Nature 433, 53–57 (2005).

  6. 6

    Driscoll, N. W. & Haug, G. H. Science 282, 436–438 (1998).

  7. 7

    Delaney, M. L. & Boyle, E. A. Paleoceanography 3, 137–156 (1988).

  8. 8

    DeConto, R. M. & Pollard, D. Nature 421, 245–249 (2003).

  9. 9

    Zachos, J. C. & Kump, L. R. Glob. Planet. Change 47, 51–66 (2005).

  10. 10

    Kump, L. R., Kasting, J. F. & Crane, R. G. The Earth System 2nd edn (Prentice-Hall, Upper Saddle River, NJ, 2004).

Download references

Author information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Kump, L. Foreshadowing the glacial era. Nature 436, 333–334 (2005) doi:10.1038/436333a

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.