Speech given to the College of Biological Science -- 6 April, 1996

Breakthroughs in Ice-Age History

When I was a student the main discussion about ice-age history was whether three or four periods of glaciation occurred, and how much time has elapsed since the last glaciation. One of the leading measures of the length of postglacial time was right here in St. Paul -- the rate of retreat of St. Anthony Falls on the Mississippi River, estimated by Winchell in 1872 to represent about 8,000 years on the basis of the retreat since 1670, as sketched by French explorers, and extrapolation of this rate back to Fort Snelling, where the falls started as a result of the catastrophic drainage of Glacial Lake Agassiz in northwestern Minnesota and adjacent Manitoba when the continental ice sheet retreated. Winchell's estimate, which later proved to be not far off, was made not many years after Darwin, when the scientific world was uncertain about geological time and evolution.

In my first field adventure when I went back to academic life after the Second World War, I joined some archaeologists in coastal Lebanon in trying to date certain cave deposits by the loose geologic/climatic methods of the day, and this led to other archaeological projects in Iraq, Iran, and Turkey, where the problem posed concerned the relation between the climatic change at the end of the ice age and the origin of agriculture-- a question raised by early archaeologists a century ago and brought into focus in the 1930s by the great English archaeologist V. Gordon Childe, who proposed that the climatic change at the end of the ice age reduced woodlands of the Near East to semi-deserts, with oases that drew nomadic peoples together with animals and edible plants and led to domestication of both. At that time the only chronology available for the Near East was archaeological, based on Egyptian king lists dating back to about 5000 years ago. At some uncertain time before that, cultures had evolved from nomadic hunting groups to city states in Mesopotamia and Egypt, supported largely by irrigation agriculture. At some time uncertain time also it was known that the global climate had changed, bringing about the wastage of continental ice sheets and shifts in patterns of vegetation, animal life, and the elements of climate.

The first task was to demonstrate contemporaneity between cultural change and climatic change. Here a major breakthrough in ice-age history came in about 1950, and it made it possible to date both types of event -- radiocarbon analysis, which depends on cosmic-ray production of the isotope carbon-14 in the atmosphere, its incorporation into plant material, and its decay at a rate that makes possible the dating of fossils back to about 40,000 years ago. Archaeologists immediately took advantage of the new method and dated plant remains from the earliest agricultural sites as about 10,000 years ago, and at the same time, paleoecologists were able to reconstruct vegetation changes and thus climatic history in the eastern Mediterranean region on the basis of pollen analysis of organic lake sediments, documenting a major shift about 10,000 years ago from cold semi-arid steppe to woodland in a strongly seasonal climate. Of course contemporaneity does not demonstrate cause and effect, but it is now possible to evaluate the cultural factors that brought about the agricultural revolution in the context of climatic change and the associated changes in the flora.

Now, 35 years after the introduction of radiocarbon dating, dozens of laboratories are set up to provide analyses, and hundreds of thousands of dates have made it possible to reconstruct natural and cultural history throughout the world. Recent developments of carbon-14 measurement by accelerator mass spectrometry rather than by decay counting has made it possible to use samples as small as a few milligrams, although the limit of about 40,000 years still pertains. At the same time one of the early assumptions of the method -- that carbon-14 production in the atmosphere has been constant -- has been shown to be slightly false, for dating of tree rings back to about 10,000 years ago shows wiggles in the carbon-14 curve, probably because of variations in cosmic radiation and in earth-magnetic factors. Major wiggles around 10,000 years ago may have been caused as well by circulation changes in the ocean, which is a major reservoir of carbon.

History is nothing without chronology. It enables one to date events in one area with those in another and thus detect geographic patterns of change. As an example on the ecological side, the vegetation history of the Minnesota area, reconstructed from pollen analysis of lake sediments, shows the replacement of spruce forest by pine forest progressing from south to north 12,000 to 9000 years ago, and of this by temperate trees leading to prairie, then followed by a reversal to the present time. Thus, it was substantially warmer and drier 7000 years ago than it is today. Such changes furnish the settings for study of the climatic, edaphic, and biological relations involved in the transformations from one major vegetation type to another.

But what of the longer time range in ice-age history? Fifty years ago, little was known, except that several periods of continental glaciation had occurred. The. next breakthrough was the result of the acquisition of deep-sea cores, which record steady sediment deposition for millions of years. The sediment contains fossils of planktonic and benthic organisms whose geographic distribution in the world oceans is controlled primarily by the water temperature and salinity. The stratigraphic abundance of planktonic microfossils of known ecological requirements in marine sediment cores permits the reconstruction of changes in the sea-surface temperatures. Dates before 40,000 years ago were made possible by the chronology of reversal of the earth's magnetic field as recorded in these sediments dated by correlation with radiometrically dated sequences of lava flows back for millions of years. But even more important than this evidence for changing sea-surface temperatures is the stable-isotopic composition of the carbonate shells of these organisms. When ocean water evaporates, the molecules containing oxygen-16 are more numerous, and the heavier oxygen-18 is left behind. As the air masses move over the continents and cool, the vapor condenses, and the rain and snow with the heavier isotope preferentially falls out. During the glacial period, the snow at high latitudes had progressively less oxygen-18, so the glacial ice that built up was isotopically light. By this mechanism, the oxygen-isotope stratigraphy of shells in marine sediment records the isotopic composition of the water left behind in the sea, and therefore it indirectly records the volume of ice sheets and glaciers.

The stratigraphic pattern of fluctuations in the isotopic composition of the progressively buried microfossil shells bears a strong resemblance to variations in the Earth/Sun orbital cycles calculated long ago by the Serbian mathematician Milankovitch -- eccentricity of the Earth's orbit, inclination of its axis, and precession of the equinoxes, in cycles of 100,000 years, 40,000 years, and 21,000 years. The isotopic fluctuations in the marine cores were particularly large during the last 700,000 years and had a frequency of about 100,000 years, implying seven times of major continental glaciation. Before that, the 40,000 year cycle with smaller amplitude predominated, and the North American ice sheet probably was restricted to Canada. Before about 2.4 million years ago, the isotopic fluctuations were much smaller, and they probably record only changes in ocean temperature rather than storage in ice sheets. This conclusion is supported by the first appearance 2.4 million years ago of detritus dropped into the ocean sediments by icebergs discharging from ice sheets that reached the sea.

The implications of the marine isotope stratigraphy for terrestrial biogeography are considerable. It appears that for at least the past 2 million years, the climate of now-temperate regions has been changing back and forth from cold to warm conditions, and that about 80% of the time it has been colder than the present. Continental glaciation will certainly come again. It's only a matter of time.

I have not been a direct participant in these explorations of global events. Rather I have been involved in gathering data in the field about events that can evaluate the effects of climatic change on the landscape. I recount some of them to give you a flavor of my approach to the history of landscapes during the ice age and later. I started in the American Southwest in a beautiful mountain area on the Navajo Reservation of New Mexico. Then for practical purposes here in Minnesota I worked out the complicated record of ice lobes coming off the big ice sheet, which produced the landforms and lakes that make this region so attractive. The area was especially interesting not only because of the complex glacial history, but because the three major vegetation formations -- conifer forest, deciduous forest, and prairie -- occur within the state, so that the history of their ecotones can tell a lot about climate history.

Meanwhile I became more involved with archaeologists in the Near East, and I spent parts of six field seasons in Iraq, Iran, and Turkey working on the whole problem of environmental reconstruction for the time of village settlement and the origin of agriculture. Enroute to the Near East, I commonly spent some time in the field in southern Greece on the comprehensive interdisciplinary archaeological project of Bill McDonald of our university, where again one of the problems was environmental reconstruction for the time of the Late Bronze Age culture.

But here at home, other facets of natural landscape development prompted some investigation. The wilderness forests of the northern Minnesota lake country were a mosaic of different tree communities, and it had been proposed that fire was an important factor in establishing and shifting this mosaic over time. This led to more studies of the forest history by pollen analysis of lake sediments, combined with charcoal counts to evaluate the frequency of fire. But just at this lime in 1972, the Little Sioux fire broke out and spread into the wilderness area, providing the opportunity for a group of ecologists and limnologists to examine the subsequent plant succession as well as the effect of the fire on the soils and the lakes. This experience of assembling scientists of different backgrounds to work together in the field led me to organize an expedition to the St. Elias Mountains in the Yukon to study the development of the landforms, vegetation, and lakes that characterized the stagnant toe of a surging glacier, features that were similar to the Minnesota landscape produced during the last glaciation. The party consisted of a glacial geologist, an ecologist, two limnologists, and a philosopher of science, whose task was to observe how these people applied different approaches to a common problem during a month in the wilderness.

My experience with forest fires led to another encounter with archaeology. Bill Fitzhugh of the Smithsonian Institution had heard about our work in the conifer forest of northern Minnesota, and he had the idea that the movement of prehistoric peoples in his field area in Labrador was related to caribou migrations, and that these in turn were affected by the incidence of forest fire, which opened the forest for growth of the plants that provided their main food resource. Fire frequency could be determined by charcoal analysis of lake sediments. I was tempted to work in another wilderness area on an interesting problem, so during the next six summers I spent in southeastern Labrador midst the black flies and mosquitoes, which are more numerous than needles on the black spruce trees that dominate the area -- a failure of evolution, I maintain, to develop adequate predators for those creatures. The project soon developed into a general study of the vegetation and ecological processes related to fire and other factors, as well as the overall vegetation history and lake development. I served principally as field assistant for the four graduate students who undertook thesis research on different facets of the problem. Besides the insects, the main hazards in field work were the bush pilots, one of whom left us for four days without food because he forgot to pick us up. Bears were also a problem, for they are attracted to bright yellow rubber boats unattended. Incidentally, the hypothesis about caribou and prehistoric people was left behind as intractable, but it had served its purpose in attracting me to a wilderness region with interesting ecological problems.

About this time, when the novelty of Labrador had worn off, another archaeologist, an acquaintance dating back to the Near Eastern days, called to ask if I could make some environmental reconstruction for the scene in the high Andes of Peru, where llama hunters had lived in eaves at 4000 m. I asked if any lakes were nearby, because lakes are the most effective archives of environmental history. He was uncertain. But I decided to go anyway, and I subsequently spent parts of six summers in the Andes working on the glacial geology and lakes, which were in fact abundant. When field work in highland Peru became unwise because of the hazards of guerrillas, we moved to Bolivia. In all this Andean work it was apparent that the glaciers were rapidly retreating. Some that are visible on 1960 air photographs have entirely disappeared. This phenomenon is well established in other mountain regions as well, like the Swiss Alps, and it is strong evidence for significant climatic warming in the last century or so.

Apart from these major efforts, I made single-year trips to Alaska, Antarctica, Australia, New Zealand, and Kenya, as well as less exotic places like Sweden, Norway, Ireland, and Switzerland, but these did not develop into main projects. I recount these experiences only to show that an academic field geologist/ecologist has opportunities to study diverse landscapes and fit them all together in the context of climate and climatic history. Of course one must have a domestic circumstance that permits such summer travels, rather than dedication to a cabin by the lake. I might add that on many of these fact-finding expeditions I was able to take along one or another of a houseful of boys as helpers -- to the Navaho country, the Canoe country of Minnesota, and to Iraq, Iran, Yukon, Labrador, Peru. In 1954 I drove with my family in a Land Rover and luggage trailer from London to Baghdad, by way of Interlaken, Graz, Belgrade, Istanbul, Damascus, and the pipeline road across the desert of Jordan, camping just off the road in cow pastures and by-ways for the entire trip for a month, for this was long before camp grounds existed.

The title of this talk, "Breakthroughs in Ice-Age History", is ambiguous, Is history a recounting and interpretation of events, or is it the events themselves? So far I have mentioned two breakthroughs in technology and methodology that greatly changed our understanding of the natural course of events over tens or thousands and even millions of years -- the development of radiocarbon dating, and the development of oxygen-isotope analysis of deep-sea sediment cores and its relation to Earth/Sun orbital cycles. Between breakthroughs, the individual scientist works along with prevailing concepts, making personal mini-breakthroughs every time a new insight evolves or a hypothesis is adequately proven or falsified. It is in those categories where fit the several interdisciplinary adventures that I have recounted. All scientists have these experiences, and that's what keeps them going.

One more breakthrough in ice-age history I'd like to mention before returning to the ambiguity of the term is the development of numerical paleoclimatic models, which involved the modification of supercomputer weather-forecast models to account for different boundary conditions for various times in the past according to well-established relations. For the last glacial period, the extent and height of the ice sheets, the sea-surface temperatures, the content of carbon dioxide in the atmosphere (according to bubbles in cores of Antarctic and Greenland ice sheets), and particularly the variations in seasonal insolation related to astronomic cycles are all well known. Model simulations for the last glacial maximum 20,000 years ago have shown that the expansion of lakes in the now-desert basins in the American Southwest resulted from a southward displacement of the main jet stream and its associated storms by the ice sheet, and in fact a splitting of the jet stream so that a branch went across the arctic region north of the ice sheet and brought cold air the North Atlantic, which helped to produce the frigid climate and tundra vegetation of Europe all the way to the Alps. Model simulations for 9000 years ago show that the increased summer insolation in the heart of Eurasia raised the summer temperature as much as 7 degrees Celsius above today's, thereby enhancing the summer monsoons as moist air was drawn in from the Indian Ocean. Lakes in the East African and southern Arabian deserts expanded at that time, and mammal and human populations expanded in response to the greater vegetation cover. Another paleoclimatic model simulation for a much earlier time 3 million years ago showed that before the mountains of western North America were uplifted, as well as the Himalayan/Tibetan region, the jet stream went around the Earth with little perturbation, but when these mountain uplifts reached a critical height, loops in the jet stream developed, favoring the southward expansion of arctic air over North America in summer and the northward expansion of moist subtropical air in the North Atlantic region. Both of these trends favored the growth and preservation of glaciers in Canada and Scandinavia -- thus the beginning of continental glaciation. An additional factor at about this time was the formation of the Isthmus of Panama, which changed inter-ocean circulation.

Now, finally, about the breakthrough in actual events. Its revelation is a result of a technological development as well. I refer to the successful drilling to the bottom of the Greenland ice sheet. One of the participants in the project gave a seminar here last year entitled "The ice age ended in 3 years". The chronology of the ice core was determined by counting annual layers back to more than 13,000 years ago, formed because oxygen-isotope composition depends on seasonal temperature of snowfall, and also the dust content detected by electrical conductivity variations is also seasonal. The stratigraphic record shows an abrupt increase in inferred atmospheric temperature over a very short period. This was truly a breakthrough in climate from a glacial mode to the present postglacial mode. Pollen analysts and other students of the terrestrial scene are scrambling to find evidence from lake sediments and other archives for the rate of response of different plant and animal groups to such an abrupt climatic change. Actually, the abruptness was a result of the fact that the climate was already getting warmer as early as 13,000 years ago as a result of increasing summer insolation and warming of the North Atlantic surface water. Glaciers were retreating far up into the Alps. Fossils show that beetles and other easily dispersed organisms like water plants were colonizing the tundra areas of Europe. But then came an abrupt episode of cooling that lasted about 1000 years. The reason is believed to be a massive discharge of icebergs into the North Atlantic from the critically wasting ice sheets. Icebergs are very effective in cooling the ocean surface because of the latent-heat effect, and the cooling dominated even though summer insolation was still increasing. When the discharge ceased and the icebergs had all melted, the North Atlantic sea surface rapidly warmed to catch up in its response to summer insolation, which had reached its maximum by this time. The Scandinavian ice sheet rapidly disappeared, the associated cold dry winds out of Siberia that had occurred at times during the cold period decreased in frequency, and the greater humidity of the climate in Europe allowed trees to expand rapidly from their ice-age refuges in the Balkan and Italian mountains. Within a very few hundred years Europe was covered by forest as continuous as today's.

Is this the kind of rapid climatic change that is store for us in the 21st century with a doubling of carbon dioxide? Is this to be another breakthrough in the climate itself, occasioned this time not by astronomic or glaciological processes but by modern industrial societies? There seems to be no doubt that global temperature and atmospheric carbon dioxide content are linked, as shown by the atmospheric archive represented by bubbles in the ice cores, as well as by such indicators as stomatal density on fossil leaves. The responses of various components of terrestrial and aquatic ecosystems to rapid climatic change are complex and partly unpredictable, because of feedbacks and interactions, but studies of the climatic breakthrough of 10,000 years ago provide some insights. It is likely that many small and large breakthroughs in our understanding of ecosystem processes will be necessary before either natural or artificial climatic breakthroughs can be fully evaluated. I close with the conviction that, despite our limitations in full understanding, the past is not only the key to the present but to the future as well.

E-mail Herb at hew@maroon.tc.umn.edu
Last updated October 22nd, 1996.

Back to Herb Wright's Home Page.