The missing snowballs

Another small detail missing from the Canyon record is any record of Precambrian glaciation. In recent decades, the idea that the late Precambrian Earth was encompassed by at least one global glacial event has become a topic of much debate. This so-called Snowball Earth theory has generated a lot of academic heat and many questions have been raised over the nature of the supporting evidence. Buckets of academic cold water are constantly being thrown on the wilder aspects of the theory.

As long ago as 1949, Sir Douglas Mawson first suggested that the occurrence of very ancient glacial deposits in the Flinders Range of South Australia indicated that a glaciation had extended into much lower latitudes than ever thought possible. Mawson also suggested that climate amelioration following déglaciation paved the way for the evolution of the first multicelled animals.

Sir Douglas Mawson, 1882-1958, English-bom University of Sydney-trained geologist in the University of Adelaide (1905-52) and explorer with Shackleton's Antarctic expedition (1907-9), Australasian Antarctic Expedition (1911-14, 1929-31), knighted 1914.

By then, Mawson knew that a protégé of his, Reg Sprigg, had recently found evidence for such animals in associated rock strata from the same region of South Australia (see below). However, Mawson believed that the position of the continents was fixed. As it became clearer that this was not so and that instead continents had moved or 'drifted' considerable distances, Mawson's ideas were dismissed. It was assumed that, when Australia was restored to its original position at the time, it would be near the pole and the ancient glacial deposits would be more easily explained.

Reginald Claude Sprigg, 1919-94, University of Adelaide-educated South Australian economic geologist and student of Mawson's, who worked for the South Australian department of mines then formed his own company in 1954. Elected youngest fellow of Royal Society of Australia at age of 17.

When in 1964 Cambridge geologist Brian Harland reviewed the growing evidence from around the world for a glaciation in late Precambrian times, he found evidence that seemed to indicate the occurrence of perhaps two separate glaciations spreading from the poles into low latitudes. The problem was that nobody knew where exactly the continents were positioned in late Precambrian times. Although there was some palaeomagnetic support for the idea, the data were not generally accepted as good enough to support such a revolutionary idea. But now we have somewhat better information, thanks to the painstaking accumulation of palaeomagnetic and radiometric data from these ancient rocks. Altogether, these data allow us to have a better idea of the placing and timing of individual continent positions relative to the poles. However, it is important to

Brian Walter Harland, 1917-2003, British, Cambridge-trained geologist who taught in China (1942-6) before returning to Cambridge. He specialised in the geology of Spitsbergen and the Arctic in general and was one of the first to claim that there had been widespread glaciation in Precambrian times.

realise that the state of our knowledge is still very incomplete and far from satisfactory. Nevertheless, there is a lot of new evidence to show that something very odd was going on with global climates at the time, even if it did not amount to global Snowball Earth conditions.

There is only one really secure, well-dated and reliable low-latitude data point and that is for the Sturtian age Elatina glacial deposits of South Australia, plus another possible one of similar age in Oman (around 723 million years old). These rocks can be very difficult to date and so much has happened to them since they were originally deposited that it can also be frustratingly difficult to establish their original positions relative to the poles of the time. The ongoing research and data gathering is all part and parcel of the general attempt to reconstruct the distribution of past continents and plate movements, along with the formation and break-up of ancient supercontinent agglomerations such as Rodinia.

The Snowball Earth theory has now grown to such an extent that some experts suspect that there might have been at least three late Proterozoic glaciations, first between 740 and 700 million years ago (called the Sturtian), followed by the Marinoan (650-630 million years ago) and the Gaskieran between 590 and 580 million years ago. In addition, there may also have been a restricted glaciation right at the end of Proterozoic times around 550 million years ago and a much older one around 2.3 billion years ago when banded iron formation was common.

The story of how a global glaciation could have happened goes like this. When an equator-spanning supercontinent like Rodinia begins to break up, tectonic uplift of land surfaces greatly increases the rate and extent of weathering and erosion. Carbon dioxide from the atmosphere mixes with rainwater to form weak carbonic acid, which reacts with freshly weathered debris of carbonate-rich rocks such as limestone, converting it to soluble bicarbonate. Not only is carbon dioxide removed from the atmosphere, but the dissolved bicarbonate is flushed into the oceans, where it combines with calcium and magnesium to reprecipitate and produce huge volumes of new carbonate sediments. These sink to the seabed and are 'locked up', forming a vast store of the Earth's carbon budget.

As carbon dioxide is a greenhouse gas, the removal of very large volumes from the atmosphere leads to significant and rapid cooling. Descent into an icehouse state initiates glaciation and the formation of ice caps. Incoming light energy from the Sun is very effectively reflected back from large ice surfaces (the albedo effect) and feeds back further cooling of the atmosphere (perhaps as low as minus 50 degrees Celsius) and growth of the icecaps. This in turn leads to runaway glaciation extending into tropical low latitudes, perhaps even encompassing the globe in a kilometre-thick layer of ice to produce a Snowball Earth. Most of the life on Earth is wiped out. But as suddenly as the glaciation formed it can come to an end.

While glaciation is doing its worst, global volcanoes still blast out large volumes of gases such as carbon dioxide into the atmosphere; even ice caps cannot prevent the process. In addition, land surface processes of weathering and erosion would have been literally frozen and put on hold. As a result, the associated chemical cycles that normally consume atmospheric carbon dioxide would have come to a halt, allowing this greenhouse gas to accumulate. Gradually the atmospheric carbon dioxide is replenished. And once a critical threshold is reached (estimated to require around 350 times present carbon dioxide levels), the atmosphere-ocean system suddenly flips from an icehouse into greenhouse or hothouse state, as some prefer to call it. Temperatures rise so rapidly (perhaps as high as 50 degrees Celsius) that within a few hundred years most of the ice melts. Sea levels rise and renewed weathering and erosion of 'de-iced' landscapes rapidly flushes bicarbonate into the oceans, quickly depositing a thick blanket of carbonate sediment, now preserved as cap carbonates on glacial deposits.

For this Snowball Earth theory to be correct, a number of features need to be observed. For instance, there should be global synchronisation of glaciations at all latitudes. This is still very difficult to prove, as many of the so-called glacial deposits cannot be reliably dated. Even with the rapidity with which glaciation and déglaciation are thought to have happened, for it to be global, it would have to be long lasting, in the order of 10 million years. Again, timing is problematic without good dating and it is now turning out that some sections have several diamictites (glacially derived sediments) interlayered with sandstone strata, which does not seem to fit the expected pattern. The radical changes in carbon dioxide levels should be reflected in the rock record, along with evidence for greatly enhanced weathering. New measures of weathering derived from independent isotope studies (of the element strontium) do not support the very high levels of weathering suggested by the model.

The main sedimentary signatures of late Proterozoic glaciations are first pebble-filled mudrocks that are called diamictites, and in this context they are interpreted as glacial in origin. Unlike deposits of the geologically recent Pleistocene ice ages, these ancient diamictites are invariably sandwiched between limestones that form the second signature. The limestone strata were deposited as carbonates characteristic of warm waters and warm climates, with virtually no intermediate sediments in between. In addition, there are some sedimentary iron formations that are making their last appearance in the rock record.

The discovery and interpretation of some late Proterozoic-age diamictites as glacial deposits was fundamental to the claim that there had been glaciation in this interval of Earth Time. For some diamictites there is no doubt about their glacial origin. They not only contain ice-scratched angular pebbles and boulders, but they are also associated with other glacial features such as glacially striated rock pavements and dropstones. The latter are pebbles and boulders that have fallen from sea ice into more normal marine sediments. However, in the rush to join the Snowball bandwagon, many other diamictic conglomeratic sandstones have been called glacial, with very little supporting evidence for the attribution. Critics of the 'full-on' Snowball theory claim that out of 85 diamictite deposits from around the world, dated at between 800 and 500 million years old, only some 16 are reliably glacial in origin.

The presence of limestones immediately adjacent to the diamictites is part of the supporting evidence for low-latitude positions of glaciated regions. Furthermore, the reappearance of limestones immediately after the diamictites seems to indicate very rapid climate change. However, the composition of the limestones is not exactly the same. Analysis of their carbon isotopes shows that there was a considerable organic contribution from sea-dwelling creatures to limestones found below the glacial deposits, but very little contribution to the limestones immediately above.

High positive carbon isotope ratios in limestones are also associated with large areas of shallow tropical seas. And these in turn are more abundant when supercontinents such as Rodinia rift and break apart. During such rifting, the total length of coastlines more than doubles as shallow seas flood around the margins of the new, smaller continental blocks or plates. The high rates of burial of organic carbon also locked up very large amounts of atmospheric carbon dioxide, contributing to the ice-house effect with its rapid and drastic cooling leading to a glacial event. Then, if the glacial event was globally extensive, the covering of sea ice would have shut down primary production in the oceans and led to the shutdown of marine ecosystems.

By contrast, the cap limestones above the diamictites generally have very low negative carbon isotope levels. At first these negative values seemed to agree with the Snowball model, in that they suggested continuing low organic productivity in the oceans, thus the organic component. Alternatively, the postglacial ocean may have continued to be depleted in the kind of organisms, such as cyanobacteria and algae, which were the source of the organic carbon.

However, recent more detailed analyses show that the lower strata of the cap limestones exhibit high positive ratios and the negative excursion does not 'kick in' until some way through the deposits. Depositional rates of the limestones could have been enhanced as a result of rapid heating of tropical waters and greatly increased biological productivity. But at the same time, increased weathering of continental rocks removes carbon dioxide from the atmosphere.

The removal of large volumes of this greenhouse gas would have led to global cooling. Descent into an icehouse state would have initiated glaciation. Another possibility that is much in vogue at the moment is that large quantities of methane, which has low carbon isotope values, were suddenly released from permafrost sediments that contain large volumes of frozen gas hydrates such as methane.

However, recent more detailed sampling and analysis of the distribution of carbon isotopic values through the carbonates show some puzzling complications. Positive values continue upwards into the base of the carbonates immediately above the glacial deposits. This strongly suggests that, contrary to what has been claimed, there was continuing organic production through the glacial phase. Furthermore, the claim that the brief negative carbon isotope 'excursion' was related to increased weathering has not been supported by analysis of strontium isotopes, which are also involved in the process.

Iron-rich sediments were last seen around 2 billion years ago in Earth Time and their reappearance at this late stage, when levels of oxygen in the atmosphere and oceans were close to those of the present, might seem rather strange. The answer given by Paul Hoffman and Dan Schrag, some of the main promoters of the Snowball Earth hypothesis, is that a covering of sea ice shut down the primary production of oxygen in the oceans by photosynthesising micro-organisms such as algae and cyanobacteria. Furthermore, they claim that the 'icing' on the oceans also prevented the diffusion of oxygen from the atmosphere into ocean water, promoting the deposition of iron-rich sediment in ocean waters.

That there were significant glacial events in late Precambrian times is not in doubt. But recent detailed examination of some of the critical evidence for their extent and influence on life does not seem to be quite clear cut as originally 'advertised'. There is certainly good evidence that there was no total ice blanket, there were probably significant 'uniced' areas of landscape and open areas of ocean, and life was not shut down. The Snowball Earth hypothesis has not melted away, but it is in the process of reinvention and transformation. So far, the Canyon deposits do not provide evidence for these events, but there is geological evidence not far away in eastern California.

0 0

Post a comment