Ice Over Earth.
Neatly displayed in exposed rocks in Wales is an extraordinary message first recognized by a prominent British geologist, Adam Sedgwick, nearly 200 years ago: a noticeably fossil-rich layer lies atop a layer of fossil-deficient rocks. Sedgwick christened the fossil-rich layer the Cambrian period and thereby laid the foundations for the geological timescale. Eventually, time embodied in the older rocks became known as the Precambrian era. Sedgwick went on to define the Paleozoic era (meaning "early life") based on grouping the Cambrian layer with several higher layers, all dominated by invertebrate fossils. Today we know that the Paleozoic spans more than 300 million years, ending with the rise of dinosaurs 225 million years ago.
We now know extensive details about the history of Earth since it formed 4.6 billion years ago. Single-cell fossils first appeared 3.5 billion years ago. Sedgwick chose the boundary between the Precambrian and Cambrian (575 to 525 million years ago) at the point in the rock record where he noticed a profound change from dominance by single- celled life to an explosion of multicellular life. Ever since Sedgwick's discovery, geologists have struggled to explain it.
At the end of the twentieth century, geologists found themselves confronted with an astonishing hypothesis. It aimed to tie together several pieces of otherwise anomalous geological data in a way that even attempted to explain the Cambrian's explosion of life. Writing in the journal Science in mid-1998, Paul Hoffman, Galen Halverson, Daniel Schrag, and Alan Kaufman, all then at Harvard University, proposed that preceding the Cambrian explosion Earth had passed through a period in which it was encased in ice.
The geological record clearly points to the glaciation, they said, and further the ice age terminated when carbon dioxide released from volcanoes created a horrendous greenhouse effect and global warming. As Earth gradually cooled, they contended, multicellular life sprang from the chaos. Gaining acceptance of such a big idea requires considerable supporting evidence and defense against counterevidence, as we shall see. First the main proposal, snowball Earth.
An idea germinates
The snowball Earth hypothesis germinated in 1964 in the mind of W. Brian Harland of the University of Cambridge when he noticed that glacial deposits (called tillites) from the Neoproterozoic era are found on rock outcroppings on every continent, with the possible exception of Antarctica. The Neoproterozoic covers more than 200 million years immediately preceding the Cambrian layer identified by Sedgwick.
Some of these deposits are thousands of meters thick and contain striated rocks thought to have been gouged by glaciers as they scraped across the land. Harland understood that the tillites were glacially formed and realized that they appeared in the rock record just below strata bearing Ediacaran fossils, the earliest abundant multicellular life detected in the rock record.
Harland suggested a relationship between the glaciation and the earliest multicellular life, but there were not enough data to verify his supposition. Geologists now know that there were two periods of extensive glaciation in the Neoproterozic: the Sturtian and Varanger glacial intervals, from approximately 780 to 700 and 610 to 575 million years ago, respectively.
Although the concept of plate tectonics was still in its infancy, Harland suspected, based on the available records of magnetic field directions frozen into the rock record, that during the Neoproterozoic the continents all clustered together near the equator. He grounded his conclusions on the knowledge that Earth's magnetic field lines are parallel with the surface at the equator but angle more and more sharply into the ground at higher latitudes approaching the poles [see "Demystifying Magnetism," The World & I , October 2000, p. 148]. Harland accumulated enough data from the scientific literature to show that in the Neoproterozoic tillites, the mineral grains susceptible to magnetism had assumed a nearly horizontal orientation when the glacial deposits were formed, a clear sign that the tillites originated near the equator. His conclusion was that large glaciers existed over at least sections of all the continents at the equator during the Neoproterozoic era.
Subsequent studies, including the discovery of a plethora of "dropstones" in equatorial sediments, verified Harland's thesis. Dropstones include large rocks freed from icebergs as they melt. The dropstones are easily recognized, because they are deposited along with the finer sediment that rivers carry into the oceans and ocean currents move to deep water.
Despite the mounting evidence, the scientific community was perplexed by the idea of equatorial glaciers and icebergs. Of course, glaciers do exist today near the equator on Mount Kenya in East Africa (0.5[inverted exclamation mark] latitude), and glacial deposits have been documented in Hawaii. These glaciers and deposits, however, occur at high elevations; they represent no threat to Earth's present climate equilibrium and no challenge to current geological and climatological models.
The ice-catastrophe climate model
In 1969, Mikhail Budyko of the Leningrad Geophysical Observatory developed the ice-catastrophe climate model. He used some of the earliest mathematical models of Earth's climate, which, though primitive by today's standards, nonetheless allowed him to predict climate changes based on the way Earth's surface conditions affect the heating impact of solar radiation. Budyko's work, which focused on the feedback system set up as ice reflects sunlight, revealed the important and complex role of carbon dioxide in regulating global temperature.
If, for example, the concentration of carbon dioxide, a greenhouse gas, declines in the atmosphere due to some natural process, the loss of warming might permit temperatures to drop low enough to allow the formation of glaciers on continents in equatorial regions. One mechanism for removing carbon dioxide from the atmosphere is intense weathering of the continents.
Such a process delivers calcium and magnesium into the seas, where the dissolved metals combine with carbon dioxide to form carbonates of calcium or calcium and magnesium along the bottom of the oceans as precipitates, thereby depleting the atmosphere's carbon dioxide content. By this process, intense weathering could cause enough cooling to trigger the formation of equatorial glaciers.
The argument continues. Snow and ice have a high albedo--that is, they extensively reflect sunlight back out into space. Thus areas covered with snow and ice tend to cool Earth in comparison to land or water areas not covered with snow or ice. Most of the Sun's light strikes equatorial latitudes (between 0 and 30[inverted exclamation mark]). Budyko found that once glaciers are born in equatorial regions, they initiate a spiraling feedback loop: The cooling from the reflected light of the snow at the equator increases glaciation, and more glaciation increases the albedo. Based on the model, he proposed that the runaway system would lead to an ice catastrophe engulfing the entire planet in glaciers, but he still doubted that Earth was ever encompassed in ice.
Budyko, like almost all the scientists of the day, assumed that an ice- encased Earth would destroy all existing life. In the 1970s, bacterial life was discovered in several places, including deep-sea vents and in rock miles below the surface (oil-rich environments), where it lived on energy supplied from Earth's interior rather than the Sun. Perhaps life could survive a deep freeze, but most scientists were still incredulous about an Earth encapsulated in ice.
Another weakness of the ice-catastrophe scenario was its failure to explain how a deep-frozen Earth could return to normal. Joseph Kirschvink of the California Institute of Technology offered a possible answer in 1992. He was the first to use the term snowball Earth. In a published article he reminded scientists that, during an ice catastrophe, plate tectonics would continue to generate volcanoes in the same fashion that the churning of the plates has done throughout Earth history. And those volcanoes would continue--just as they do today--to produce large quantities of greenhouse gases, in particular, carbon dioxide.
Further, ice covering the land and oceans would halt the hydrologic cycling of water, including the rain necessary for eroding the continents. With no mechanism for removing carbon dioxide from the atmosphere, it would accumulate, raising Earth's temperatures to such extraordinary levels that the ice would melt.
Understanding a key supporting argument of Kirschvink's hypothesis requires that we understand the origin of banded iron formations (BIF). These commercially valuable iron ore deposits date from 2.5 to 1.8 billion years ago and cover hundreds or even thousands of kilometers on virtually every continent. Most scientists accept the hypothesis that BIF (consisting mostly of iron oxide) precipitated out of the oceans as ocean-dwellling bacteria pumped prodigious quantities of oxygen into the water and air. In other words, Earth literally rusted. After that major rusting episode was finished about 1.8 billion years ago, the rock record began to accumulate animal cells--those cells that use free oxygen to obtain energy.
More recent, iron formations are extremely rare and smaller in scale. Kirschvink proposed that smaller iron formations interlayered with the Neoproterozoic tillites tell an important part of the snowball Earth story. In this scenario, dissolved iron levels rose when ice covering the water led to oxygen depletion in oceans, as underwater volcanoes and hot springs poured iron into the water. Then, once the ice melted, oxygen again became abundant and the iron precipitated out of the water.
Kirschvink had found a way around the twin problems of freeing Earth from the ice and forming the iron deposits in the tillites. But scientists were still hesitant to accept such a controversial idea.
The tide turns
Only after the Harvard team published its paper in Science did the tide turn. Taking the rock record in Namibia, Africa, as their primary reference, they focused their story on the carbonate rocks that almost always lie atop the Neoproterozoic tillites. These "cap carbonates" indicate precipitation in warm, shallow oceans--a stark contrast to the glacial deposits below. (Similar carbonates are currently being formed along the Bahama banks in the central Atlantic Ocean.)
The Harvard team showed that if the global warming proposed by Kirschvink occurred after the Neoproterozoic glaciation, then carbonates atop the Neoproterozoic tillites are precisely what would be expected. As the glaciers receded with the warming temperatures, the hydrologic cycle would begin to produce copious quantities of rain, particularly within the equatorial latitudes. The erosional products would gradually build up in the sea, leading to carbonate precipitation.
In support of their hypothesis, they produced some convincing physical and geochemical evidence, such as rock structures within the cap carbonates that require extremely rapid formation, probably within thousands of years. One prime example found in the Namibian cap carbonates is crystal clusters up to six feet long of the carbonate mineral aragonite. Such phenomenally large structures could only form during short intervals from ocean water extremely saturated in calcium carbonate.
Important evidence is contained in the ratios of the carbon 12 and 13 isotopes in the cap carbonates. (Carbon 12 has six protons and six neutrons in its nucleus while the carbon 13 nucleus has one more neutron.) Measuring the ratio of the two isotopes in a material is a standard procedure today, and the ratios can sometimes reveal a hidden story. For snowball Earth, the telling fact is that 99 percent of the carbon emitted from volcanoes is carbon 12 and approximately 1 percent is carbon 13.
As algae and bacteria grow in warm water, they preferentially remove the lighter carbon isotope (carbon 12) from seawater, depleting it of carbon 12 and increasing the relative abundance of carbon 13. As a consequence, carbonates precipitated from seawater that is rich in algae and bacteria will have proportionately more carbon 13 than the carbon dioxide expelled from volcanoes.
The Harvard scientists found the cap carbonates to be much lower in carbon 13 than other carbonates in the geologic record, including those being produced now. It meant only one thing to them: The period following Neoproterozoic glaciation was a time of diminished life, at least life affecting the carbon 13 concentrations. But the fact that they found carbon 13 eventually springing back to normal levels above the tillites suggested that life must have gradually increased in the warm equatorial waters. In other words, life had survived snowball Earth and had exploded into a vast variety of multicellular forms shortly after.
The Harvard team's proposal and research have taken a previously smoldering concept and ignited it into a fierce debate. In fact, their work seems to raise even more questions than it answers. Did glaciation encompass the entire planet? One group of incredulous scientists contends that glaciers could not have done so. Other problems have surfaced. Sophisticated computer climate models fail to reproduce Budyko's ice-catastrophe model. Why? How did life survive the glaciation? Can volcanoes really supply enough greenhouse gases to generate a greenhouse warming after the ice has encompassed Earth? Wrong or right, the Harvard team has certainly enlivened a field of science not noted for the rapid pace of its proceedings.
What happened to life?
So how did life survive the freeze-thaw world of the Neoproterozoic? The Harvard team suggested that as the glaciers covered Earth, life managed to survive around ocean-floor vents where hot, mineral-rich waters are supplied by magmas below. Many such communities exist today, but they are not photosynthetic. So questions remain about how photosynthetic life could have survived if Earth were blanketed with thick sheets of ice. Also, if the oceans were depleted in oxygen when Earth was covered in ice, as the iron formations suggest, animal cells should have had problems surviving even around the vents.
Looking at current evidence, Warwick Vincent of the Universite Laval in Quebec and Clive Howard-Williams of the National Insitutute of Water and Atmosphere Research Ltd. in New Zealand proposed that photosynthetic life could have survived within the ice. A summary of their work published last year in Science describes microbial mats that live on the surface of the large continental glaciers of Antarctica.
Although the mats are literally frozen in the ice, they thaw out briefly in the late summer and participate in a brief episode (about two weeks) of photosynthetic activity even though the air temperatures are below freezing. The mats, similar to others found in the Neoproterozoic fossil record, also provide "safe" habitats (protection against both extreme cold and ultraviolet radiation) for other organisms including bacteria, protists, and metazoa.
Perhaps the strongest alternative hypothesis to snowball Earth has been proposed by Mark Chandler and Linda Sohl, both at Columbia University. They used some of the world's most sophisticated computer climate models in an attempt to re-create the conditions of the Neoproterozoic. In a paper published last year, they reported that they could not find a way to cover Earth in ice even though they varied several factors during the computer runs.
Their model runs used the Neoproterozoic's pattern of continental mass distribution and included the ice-albedo effect and the luminosity of the Sun, which was 6 percent less than today. They also varied the amount of carbon dioxide in the atmosphere and the amount of heat transported to higher latitudes by ocean currents. Even starting the model run with the most extreme conditions, however, produced only 68 percent ice cover of Earth.
In fairness to the Harvard team, they noted, "These results do not necessarily preclude a 'snowball' Earth climate scenario. ... However, either more severe forcings [extremes of factors represented as parameters in their climate model] existed or radical changes occurred in the ocean/atmosphere system which are unaccounted for [by their climate model]."
Chandler told The World & I that they are currently running models that include additional factors, including changes in Earth's rotational rate, mountains on the continents, and ice already present on land. "Our results have shown that ice-covered continents were a possibility (even at low latitudes), but that ice-covered oceans were not likely to have existed for long periods, if at all, during the Late Neoproterozoic."
The work of the climate modelers has sparked a change in thinking. Some scientists have taken to referring to the snowball Earth hypothesis as the slushball Earth hypothesis. Most scientists are content with the existence of glaciers at low latitudes on land or even covering large portions of the oceans. Still, Chandler and Sohl argued persuasively that "under snowball Earth conditions, glaciation would be impossible, since the hydrological cycle would all but cease if the atmosphere's primary moisture source were cut off."
The shift in thinking toward a partially covered Earth also allows for an easier explanation of how life survived the deep freeze. The equatorial waters were free of ice and exposed to sunlight. One problem the Harvard team emphasizes as an argument against slushball Earth is that it would be difficult to concentrate iron in the oceans if the oceans were exposed to the atmosphere. The word is still out on this argument.
The ice era ends
Working on possible explanations of how the freeze-thaw Earth era ended, Martin Kennedy of the University of California, Riverside, and Nicholas Christie-Blick and Sohl of Columbia University published an exciting new hypothesis this year in Geology. They questioned the hypothesis that the cap carbonates deposited over the tillites resulted from the reaction of atmospheric carbon dioxide released from volcanoes with calcium and magnesium washed into the seas by weathering of the continents.
The rapid continental weathering required by this model should have rapidly increased the concentration in seawater of the isotope strontium 87, a signature isotope in continental rocks. But no evidence of elevated strontium 87 levels has been found in the cap carbonates, even though several studies have been conducted.
The Kennedy team has proposed instead that the cap carbonates are the result of destabilization of gas hydrates. Composed primarily of methane and water, gas hydrates are suspended in sediments as an icelike substance. These deposits are apparently so extensive that they are being seriously considered as a potential hydrocarbon energy source. Methane is produced from the breakdown of organic material buried in sediments and becomes concentrated near the sediment surface wherever temperature and pressure conditions are appropriate [see "The Ice That Burns," The World & I , June 1999, p. 162]. Permafrost regions make some of the best reservoirs.
There must have been extensive permafrost regions toward the end of the Neoproterozoic: more abundant than at any other time due to the colder conditions. The Kennedy team has proposed that gas hydrates seeped to the surface during the later part of the Neoproterozoic, gradually releasing extensive amounts of methane into the atmosphere.
They contend that methane (a greenhouse gas that is more than 20--30 times as effective as carbon dioxide) warmed the planet as it increased in the atmosphere and brought on the melting of glaciers. The team believes that the carbon contained in methane is the ultimate source of carbon in the cap carbonates.
The Kennedy group showed that structures and fabrics in the cap carbonates are similar to those found in cold methane seeps. Perhaps the most interesting aspect of their research is the fact that gas hydrates are low in carbon 13. If the cap carbonates formed from the gas hydrates, it could explain the sharp decrease in carbon 13 the Harvard scientists found in cap carbonates. In effect, the carbon 13 would have nothing to do with the absence of life that they postulated.
Do Earth's freeze-thaw cycles at the end of the Neoproterozoic explain the mystery of why the varieties of multicellular life exploded about that time? As the Harvard team has pointed out, after the Neoproterozoic glaciation--and virtually overnight in geological terms- -animals fork into the 11 distinct lineages that still exist today. They postulated that the glacial period was a time when the animal lineages were "pruned" into distinctive single-cell forms. Those surviving the glacial holocaust radiated into the varieties of animal life found in the Cambrian fossil record.
The Kennedy team begs to differ. It believes that since the carbon 13 variations may not be due to variations in the life of the oceans, we cannot say definitively that life was diminished during glaciation. In the end, we may still be a long way from understanding this mystery of life, but the termination of glaciation and the rise of multicellular life seem too fortuitous not to be related in some fashion.n
Additional Reading:Mark Chandler and Linda Sohl, "Climate Forcings and the Initiation of Low-Latitude Ice Sheets During the Neoproterozoic Varanger Glacial Interval," Journal of Geophysical Research, vol. 105, 20,737--56.
Paul F. Hoffman and Daniel P. Schrag, "Snowball Earth," Scientific American, vol. 282, Jan. 2000, 68--75.
Paul F. Hoffman Alan J. Kaufman, Galen P. Halverson, and Daniel P. Schrag, "A Neoproterozoic Snowball Earth," Science, vol. 281, 1,342-- 46.
Martin Kennedy, "Are Proterozoic Cap Carbonates and Isotopic Excursions a Record of Gas Hydrate Destabilization Following Earth's Coldest Intervals?" Geology, vol. 29, 443--46.
Marc J. Defant is professor of geology at the University of South Florida. He specializes in the study of volcanoes and is the author of Voyage of Discovery: From the Big Bang to the Ice Ages, a panoramic history of the universe, including our galaxy, solar system, and planet.
|Printer friendly Cite/link Email Feedback|
|Author:||Defant, Marc J.|
|Publication:||World and I|
|Date:||Nov 1, 2001|
|Previous Article:||The Tormented Fatherland - Wolfgang Koeppen's profound novels from the 1950s shed a harsh light on Germany's angst following World War II--and they...|
|Next Article:||The Middle East: War or Peace?|