Episode 379 Gondwana Glaciation

Today’s episode, number 379, is about glaciers in what is now the Sahara Desert, and we’re going back 340 million years, to the Mississippian or Early Carboniferous Period of the Paleozoic Era.

In the original daily episodes of this podcast back in 2014, when we got to the Mississippian in June, I had a very brief episode about glaciers in Australia and South America. They probably represented an early pulse of the well-known later glacial period during the Permian, and the glaciation provided evidence for the existence of the supercontinent of Gondwana, which was situated more or less over the south pole at that time, about 340 million years ago. 

Gondwana consisted of the continents and smaller blocks we know today as South America, Africa, Arabia, Antarctica, India, and Australia, and the pole was located somewhere in south-central Africa, so it should come as no surprise that some recent work by Daniel Le Heron at the University of London, published in the journal Geology in November 2017, reports evidence for an extensive ice sheet in what is now northern Chad, in the modern Sahara Desert.

Permian Gondwana reconstruction and inferred ice cap (blue outline) from A. du Toit, South African geologist, 1937 (Our Wandering Continents). Du Toit's work was remarkably prescient. The map above differs only slightly from modern reconstructions of Gondwana. The red oval shows the general area where Le Heron (2017) infers a small ice sheet in Carboniferous time. 

Le Heron described belts of sinuous lineaments carved by ice in belts five to 12 kilometers wide and spread over an area of more than 6,000 square kilometers. The features cut into an ancient surface that is interpreted to represent the landscape over which the glaciers flowed during that Early Carboniferous time. He suggests that the ice sheet extended to the west to cover much of what is now Niger, and that the glaciers flowed northward into the sea. The coastline then was also in far northern Chad, so the glaciers probably reached the ocean.

The present-day surface of the earth in northern Chad is actually an exhumed, an uncovered, ancient landscape that formed about 340 million years ago when those glaciers scoured the land. The surface was buried by later sediments which have since been eroded away.

I don’t think the causes for this glacial episode are well understood, although there’s been a lot of work done on it. One of the most recent reports, by Yves Goddéris & Yannick Donnadieu and their colleagues writing in Nature Geoscience in April 2017, suggests that the onset of glaciation was the result of tectonic activity. Just before the ice age got going, the Hercynian Mountains had been uplifted because of continental collisions in many parts of the world. As soon as mountains are uplifted, they begin to erode, and a lot of erosion tends to reduce atmospheric carbon dioxide because there’s a lot more material, the sediment, to react with the CO2. Goddéris and his co-workers think CO2 levels fell enough to trigger the formation of glaciers in the mountains, and ultimately by Permian time, to form extensive ice sheets that covered a huge area of Gondwana.

According to their model, the glacial period ended when the mountains had been eroded enough to affect the carbon cycle and allow CO2 levels (and overall temperatures) to rise again. The end of glaciation was also likely affected by the final amalgamation of the supercontinent of Pangaea, which changed climates to more extensive arid conditions.

If you are interested in more about the Mississippian or Early Carboniferous Period, search the archives for June 2014. All 30 episodes that month were about the Mississippian.

—Richard I. Gibson

Le Heron paper 

Goddéris paper 

Episode 373. A walk to Branham Lakes

Upper Branham Lake

Today’s episode will be a little different from what you are used to. I’m going to try to give some of the story of the Precambrian here in southwestern Montana, but I’ll do it in the context of a little hike I did yesterday to the Branham Lakes in the Tobacco Root Mountains. So there will be some of the usual narration, but also some snippets that I recorded while I was on the walk, which are not included in the script below. You can expect some huffing and puffing. See also this blog post by Pat Munday.

probably hypersthene (Mg Fe silicate)

When I was learning the geology of this region back in 1969, the Precambrian rocks of the Tobacco Root Mountains were considered to be Archean, older than 2.5 billion years. They were (and are) the northwestern-most corner of the Wyoming Craton, one of the ancient, fundamental building blocks of North America that we talked about last year. And the Wyoming Craton is definitely Archean in age. At least most of it is.

More recent analyses of age dates in southwestern Montana gave rise to another interpretation, by Tekla Harms and her colleagues a few years ago, that the zone through the Tobacco Roots, Highland Mountains south of Butte, and Ruby Range east of Dillon, Montana, represents the old margin of the craton, where a pile of sedimentary rocks formed – possibly during Archean time, but if it was then, it wasn’t long before the 2.5-billion-year cutoff date for the Archean. The sediments might have been early Proterozoic, called Paleoproterozoic. In any case, Harms and colleagues interpret age dates in some of these rocks at about 1.75 to 1.9 billion years to represent the collision between the northwestern corner of the Wyoming Province and another terrane, now mostly in the subsurface of central Montana. There isn’t much doubt that such a collision happened, but there remain questions as to whether the Precambrian metamorphic rocks of southwestern Montana were already there, Archean, or if they were sedimentary rocks that got caught up in that collision and metamorphosed a few hundred million years after they were deposited.

Geologic Map of part of the Tobacco Root Mountains. Reds and oranges are igneous rocks of the Tobacco Root Batholith, about 75 million years old. Grays are Precambrian rocks, about 1700 to 2500 million years old. Both maps from Vuke et al., 2014, Geologic Map of the Bozeman quad, Montana Bureau of Mines and Geology Open-file map 648. Black box in lower left corner is enlarged below. 

Oranges (Khto) are Tobacco Root Batholith, grays are Precambrian. X=Paleoproterozoic, about 1.7 to 1.9 billion years old; A = Archean, over 2.5 billion. XA means we aren't really sure. qfg = quartzofeldspathic gneiss, ah = amphibolite and hornblende gneiss. Xsp = Spuhler Peak formation. Branham lakes are blue. 

There isn’t much doubt that the metamorphic rocks there were originally mostly sedimentary rocks, sandstones, shales, siltstones, maybe even a few limestones, and that they were intruded by some igneous rocks like basalt, all before they were metamorphosed. We can infer what these protoliths, the original rocks, were, from the chemistry and mineralogy of the rocks today. So it’s a question that doesn’t matter too much, although it has big implications for the detailed story of this part of the world – when were sediments laid down, when were they metamorphosed. That in turn has implications for the structural and tectonic history, and understanding THAT helps us explore for mineral resources and understand things like earthquake fault distributions.

I’m not going to solve the question by walking up to the Branham Lakes. This beautiful location is about 9 miles or so up Mill Creek, east from Sheridan, Montana.

Most of the major valleys on the flanks of the mountains of southwest Montana held glaciers during the most recent glacial period that ended about 12,000 years ago or so.

Kyanite, Aluminum Silicate

Sediments like silts and muds usually contain plenty of silica, fine-grained quartz, but often they have fragments of feldspars or the clays that weather from feldspar, and those minerals contain a lot of aluminum. Under metamorphic conditions, high temperatures and pressures, the aluminum and other chemicals reorganize into minerals that are stable at those temperatures and pressures. There are three minerals, kyanite, andalusite, and sillimanite, which are chemically identical aluminum silicates, Al2SiO5, but which have different crystallography that reflects the details of the pressure-temperature regime in which the aluminum and silica were mobilized. Kyanite is stable at relatively low temperatures, 200 to 700°C, and low to high pressures; Andalusite forms at low pressures and medium temperatures, and Sillimanite forms at high temperatures across a range of pressures. The boundaries between these phases are well known so we can use their occurrence to infer the temperatures and pressures that the rocks reached during metamorphism.

Tobacco Root Batholith granite with dark xenolith of older rock

The Archean and early Proterozoic metamorphic events, about 2.5 billion to maybe 1.7 billion years ago, were ancient when the next potential metamorphic event took place, about 76 million years ago.

In the next clip, I made a mistake – I say epidote when I meant to say enstatite. They both start with an E, that’s my excuse! Enstatite is magnesium silicate, and hypersthene, also mentioned in the next clip, is enstatite with iron in it. Both are the kinds of minerals you can get by metamorphosing rocks that have a lot of iron and magnesium, probably NOT simple sediments like shale.

The road to the Branham Lakes, about 9 miles from Sheridan, Montana, is pretty good, and you could probably make it almost all the way in a 2-wheel-drive vehicle if you have decent clearance. I chose to leave my Prius about 2½ miles from the lakes just to be safe, as there are a few dicey stretches, and because it was such a fine day I really preferred to walk. If you go, it would be an unusual year that you’d find the road and lakes snow-free before late June at the earliest, but the setting is spectacular in July and August. I have a few photos from my walk on the blog, history of the earth calendar dot blogspot dot com.

I hope you have enjoyed this little ramble from the Precambrian to the Cretaceous to the glacial period of the Pleistocene. Thanks for listening!

Lower Branham Lake

—Richard I. Gibson

More photos on Facebook

December 29. The Ends of the Ice Age

I know that I’ve implied that the change from the Pleistocene glacial period to the warmer Holocene was quite abrupt, about 10 to 12 thousand years ago. And it was, generally speaking, but it wasn’t a particularly smooth change.

Dryas octopetala, photo by Jörg Hempel,
used under Creative Commons license.

Toward the end of the glacial time, as the continental ice sheets were melting back quite rapidly, various things happened to tweak the climate from one that was warming to one that was cooling again. Three of these cooling episodes are called the Dryas – Younger Dryas, Older Dryas, and Oldest Dryas. The Dryas is an Alpine and tundra-loving shrub of the rose family, the national flower of Iceland, which typifies these cool periods.

The peak of glaciation, with glaciers as far south as the Ohio and Missouri Rivers in North America and covering the British Isles in Europe, was about 21,000 or 22,000 years ago. The warming and melting that began by about 20,000 years ago was interrupted by the Oldest Dryas interval, which lasted from about 18,000 to 14,700 years ago. It appears to mirror the overall trends of the ice ages – a gradual fall in temperatures to a low point, followed by a relatively abrupt warm up over a short time span. The temperature estimates for all these events are based largely on measurements of oxygen, nitrogen, and argon ratios, which are proportional to temperature, from gases trapped in ice in Greenland and Antarctica, but they are supported by other lines of evidence too.

During each of the Dryas periods, much of Europe was tundra or taiga – Arctic conditions, but that does not mean lifeless. The taiga or boreal forest is one of the largest biomes on earth today, supporting vast forests and wide diversity of large animals, from caribou and yaks to bears and many birds. The treeless tundra is less biodiverse, but still not really barren.

After a fairly short warming period, fewer than 1,000 years, the Older Dryas cooling took place for a short time, from about 14,100 to 13,900 years ago, only a couple centuries. Its expression is largely European, so the changes may not have been global in scope.

The Younger Dryas is the best-known of these cool periods. It lasted from about 12,800 until about 11,570 years ago. It seems to have ended in a step-wise manner, in increments of 5 or 10 years over as short a period as 50 or so years. The end of the Younger Dryas is dated by various means quite accurately, to between 11,545 and 11,640 years ago, with 11,570 a common estimate.

The Younger Dryas, like the earlier events, was felt most strongly in Europe, though there is evidence for it in the Pacific Northwest of the United States. Scandinavia and Finland were under ice sheets – still, or again. Britain was largely tundra or taiga, as was most of what is now the North Sea, which was dry land supporting an extensive flora and fauna.

By now, you can probably guess at some of the speculated causes for the Younger Dryas. It’s been suggested that there was some impact at about 12,900 years ago that initiated the cool period, but I think that idea has been largely discredited. There was a decent-sized eruption of a volcano at Laacher See, near Koblenz in Germany, also at about 12,900 years ago. It was comparable to the eruption of Mt. Pinatubo in 1991, and while it may have had some effects, it’s pretty hard to see it as THE single cause of a 1,300-year cooling event.

I think the most likely cause is some change in the fundamental heat engines of the Northern Hemisphere. The focus of this line of reasoning is the circulation of warm waters to the north in the Atlantic Ocean – specifically, the Gulf Stream and the more important deep-water exchange that keeps the North Atlantic warmer. This works because of the variable density of sea water at different temperatures, so it sets up a continuous cycle of circulation and exchange.

For the Younger Dryas, the idea is that this circulation was shut down because of an influx of fresh water to the North Atlantic. This isn’t really unreasonable. The huge continental ice sheets of North America were melting, and the water had to go somewhere, but it’s a little more challenging to explain the abruptness of the changes. But we have a likely smoking gun. I talked about some of the glacial meltwater lakes in North America the other day – but I left out one of the largest – Glacial Lake Agassiz. Named for the eminent glacial geologist Louis Agassiz, this lake fronted the retreating ice sheet in what are now Manitoba, Saskatchewan, North Dakota, and Minnesota, with a possible connection to a similar glacier-margin lake that covered much of northern Ontario. The surface area was much larger than today’s Great Lakes.

Lake Agassiz map by Warren Upham, USGS Monograph 25, 1895 (public domain). The extent of the lake was actually larger than shown here.

While the continental ice sheet was still present to the north and east, Lake Agassiz drained to the south, through valleys now occupied by the Minnesota and upper Mississippi Rivers. When the ice melted enough, there could have been an emptying – either catastrophic or not, into Hudson Bay and thence into the North Atlantic. This is the influx of fresh water that is the most likely culprit in the shutdown of the North Atlantic circulation, and the cause of the Younger Dryas.

There’s one more cool period to mention – the Little Ice Age. It wasn’t really an ice age, but it was a distinctly cooler time, approximately 500 years, from 1350 A.D. until about 1850 A.D. There were several pulses of cold during this interval, well documented historically, including between about 1460 and 1550, from 1650 to 1715, and from 1770 until 1820 – and now I’m not using years ago, but the actual dates, A.D.

The likely causes are the usual suspects. One interesting one is changes in the sun’s output. Two of the coldest times coincide with periods when the sun had virtually no sunspot activity. The best known of those is the Maunder Minimum, from 1645 to 1715. Volcanic activity is of course another possibility. The eruption of Tambora, in Indonesia in 1815 famously caused the “Year Without a Summer” in 1816, and there were other major eruptions in the early 1800s, which could have impacted that cold period.

But it’s also possible that that slowdown of the North Atlantic circulation was a factor, if not THE factor. The Little Ice Age follows the Medieval Warm Period, which lasted from 950 to 1250 A.D. It was a time when the Vikings colonized Iceland and Greenland and northern Newfoundland; the latter two colonies were abandoned after the Little Ice Age began. The warm period could have caused more melting, but a more likely possibility is that it affected atmospheric circulation patterns, resulting in a persistent jet stream that might have kept Europe, especially, cooler, and might have had even global implications. For much more on this idea, and the Little Ice Age in general, I strongly recommend a book, The Little Ice Age: How Climate Made History, 1300-1850, by Brian Fagan (Basic Books, 2001)  

And for a good look at how we understand recent climate changes, I recommend The Two-Mile Time Machine: Ice Cores, Abrupt Climate Change, and Our Future, by Richard Alley (Princeton University Press, 2002). 

* * *

Before I close today I want to thank my friends, colleagues, and listeners for their suggestions, but in particular I thank geologists Patricia Dickerson in Texas, Stephen Henderson in Georgia, and Colleen Elliott right here in Butte, Montana, for their support and suggestions for topics in this series. Thanks!

—Richard I. Gibson

Younger Dryas causes 

More Younger Dryas causes 

Dryas octopetala, photo by Jörg Hempel, used under Creative Commons license.

Lake Agassiz map by Warren Upham, USGS Monograph 25, 1895 (public domain). The extent of the lake was actually larger than shown here.

December 27. Glacial and Pluvial Lakes

Today’s glacial topic is lakes again – but this time, it’s lakes that are gone. Glaciers were constantly melting, and the runoff, from on top of them, beneath them, along them, and in front of them had to go somewhere. Often, it just ran in rivers away from the glaciers, and the sediment carried by that flow created extensive deposits called outwash, glacially-derived sediment plains and mounds and more.

But depending on the topography around a glacial front, you might get a lake pooling there, especially at times when glaciers were retreating. As I mentioned yesterday, the terminus of a stagnant glacier could result in a pile of sediment called a moraine, and combined with pre-existing topography, could easily form a lake. This happened repeatedly, and we recognize lakes by their typically thin-bedded sedimentary layers. In some cases, the layers represent annual sequences like tree rings – a winter layer of minimal sediment and a summer layer when more melting brought more sediment into the lake. 

Yosemite Valley, California, is one of the best examples of a glacially-carved valley anywhere. It has the characteristic U-shape, with steep walls ground away by glacial ice. Yosemite Valley contained more than 3,000 feet of ice at the peak of the most recent glaciation. As the transition to the present interglacial period happened, probably 11,000 to 9,000 years ago, the valley floor became the site of a lake about six miles long, with the terminal moraine at the valley mouth impounding it. The situation was much like that of the Finger Lakes in New York, except in high mountain country where the bedrock is the granitic rocks of the Sierra Nevada Batholith. So the floor of Yosemite Valley today is quite flat because of those lake sediments – it’s not really truly U-shaped.


You don’t have to have a moraine to dam a river. A tongue of ice can do it, too. That’s what happened repeatedly in northern Idaho about 13,000 to 15,000 years ago, where an ice dam 2,500 feet high impounded the ancestral Clark Fork River. The resulting Glacial Lake Missoula was 2,000 feet deep and more than 225 miles long, in many branches into the mountain valleys of Montana.   

From time to time, with the changes in the climate and the build-up of pressure behind the ice dam, the ice would melt and erode until the lake began to empty catastrophically. The water volume is estimated at 60 times the discharge of the Amazon River, draining the entire lake in days to weeks. All that water poured out over eastern Washington state, where it carved the landscape into what’s called the Channeled Scablands today. There are features all over this country reflecting these events, from giant ripple marks at Camas Prairie, Montana, to dry waterfalls in Washington and wave-cut shorelines in the hills around Missoula, Montana that represent multiple lake levels. There were many fillings and emptyings of Glacial Lake Missoula, probably at least 40 times. 

There’s lots more to this fascinating story, of course, and lots to see in the area where it happened. I think your best source for detailed information about it is a book by David Alt titled Glacial Lake Missoula and its Humongous Floods (Mountain Press, 2001).

Glacial and Pluvial Lakes (and route of Bonneville and Missoula Floods)
Map by Fallschirmjäger, used under Creative Commons license 

There are a couple more big lakes that developed in the west. We call these pluvial lakes, meaning rain-derived, because they were not the direct result of snow and ice melt. Much of western Utah was covered by Lake Bonneville. The Great Salt Lake is a remnant of this lake, but Lake Bonneville covered more than 10 times the area of the Great Salt Lake. The Wasatch Mountains at Salt Lake City show two prominent horizontal lines which are shorelines of Lake Bonneville. It was more than 1,000 feet deep, and like Glacial Lake Missoula, it had a catastrophic emptying. The Bonneville Flood was about 14,500 years ago when about a third of Lake Bonneville emptied through Red Rock Pass in Idaho, onto the Snake River Plain. The dam that broke to release this flood wasn’t ice, but was topographic highs composed of lava and sediment eroded from nearby mountains.

There were many smaller pluvial lakes across the west, but the second-largest, after Bonneville, was Lake Lahontan in western Nevada. Pyramid Lake is a remnant of that lake.

The glacial period had consequences even away from the huge continental ice sheets of North America and Europe. In Bolivia, the Salar de Uyuni is a huge salt flat, covering more than 4,000 square miles or 10,000 square kilometers. It was a lake during the Pleistocene which has since dried up completely, leaving the salts behind that precipitated from the evaporating water.

* * *

On December 27, 1939, an earthquake hit Erzincan in eastern Turkey. The death toll was about 33,000, complicated by cold weather and blizzards. The quake was on or near the North Anatolian Fault, a strike-slip fault similar to the San Andreas Fault in California. The ultimate push comes from the Arabian Plate sliding northwards and forcing the small Anatolian Plate to be squeezed to the west. The Anatolian Plate is one of the blocks that made up the Cimmeride Continent, and today it makes up most of Turkey. The North Anatolian Fault separates that block from a narrow strip along the southern coast of the Black Sea, which is moving to the east relative to the Anatolian Block. Since the Erzincan quake in 1939, there have been seven strong earthquakes on the North Anatolian Fault, occurring progressively to the west. It may be that each quake adds stress to the next segment to the west, essentially triggering the later quakes.

—Richard I. Gibson

Links:
Glacial Lake Missoula
Glacial Lake Missoula video 
Salar de Uyuni 
Map by Fallschirmjäger, used under Creative Commons license 

December 26. The Great Lakes

from NASA

The origin of the Great Lakes goes back hundreds of millions of years, to the Silurian and Ordovician Periods of the Paleozoic and even longer ago. You may recall the resistant Silurian rocks that surround the Michigan Basin – but today, the important rocks are the underlying strata, the less resistant beds.

Lake Superior is underlain by the Mid-Continent Rift System, established more than a billion years ago (we talked about it January 26 and 27). There’s another fundamental break, called the St. Lawrence Rift, that formed about 570 million years ago in very late Precambrian time. It underlies the St. Lawrence River valley in Canada, where it is still seismically active, and it may extend as a weak zone southwest to where Lakes Ontario and Erie are today. And finally, around the Lower Peninsula of Michigan, alternating layers of high and low resistance were laid down over Paleozoic time, and warped into the bowl-shaped Michigan Basin.  All these things gave glaciers some paths to follow that were easier than others. 


The Pleistocene continental ice sheets had several points of origin. In eastern Canada the Laurentide ice sheet was centered in northern Quebec, but the ice flowed thousands of miles to the south and west. When it reached the area of today’s Great Lakes, it went into the low-lying areas underlain by weaker rocks, digging those areas into basins where ice thickness was somewhat greater. The ice covered everything around Michigan, including the resistant Silurian rocks, but it eroded more in the weaker zones. There’s one little area in Wisconsin where two of the main lobes of the ice sheet came together but left a small zone ice-free, but completely surrounded by ice.  

from US Army Corps of Engineers

Glaciers came and went multiple times. When you hear about glaciers “retreating,” don’t think of it as backing up. With minor exceptions, glaciers don’t back up. Glacial retreat just means that the rate of melting at the ice front exceeds the rate at which the ice is advancing. In the Great Lakes area, advances and retreats eventually made the old basins noticeably lower than the surrounding highlands, and those lowlands were filled with glacial meltwater.

The modern lakes are fed by precipitation including snowmelt, as well as by the rivers in the region. They are the largest volume of surface fresh water on earth, accounting for about 21% of the total.

While we’re on the topic of lakes, the Finger Lakes of New York also owe their origin to the Pleistocene glaciation. It’s basically the same thing as the Great Lakes, but on a smaller scale. Pre-existing river valleys were widened and sculpted by valley glaciers, and natural dams were built that created the Finger Lakes. Those natural dams are called moraines, and they represent a place where a glacier was neither advancing nor retreating, but in a continuous state of melting in place. The ice carried with it all sorts of debris – rocks and soil – which was deposited into a mound called a moraine. When that pile was at the terminus of a long valley glacier, it ended up serving as a dam to create a lake in the valley where the glacier used to be.

* * *

Today’s anniversary is one that many of you probably recall – the earthquake and ensuing tsunami on December 26, 2004, which killed an estimated 230,000 people in 14 countries. The earthquake was on the subduction zone off the northern coast of Sumatra, in Indonesia, where the Indian-Australian Plate, which is oceanic in that area, is subducting beneath an extension of the Eurasian Plate into southeast Asia and Indonesia. Most of the western and southern islands of Indonesia are volcanic, the magmatic arc that forms above the subduction zone. It’s complicated somewhat by the larger islands, especially Sumatra, having elements of more rigid crust, possibly even some small continental blocks. The quake’s magnitude of about 9.2 is the third largest ever recorded. More than 1,600 km, 1,000 miles, of fault broke and moved an estimated 15 meters, or 50 feet. That’s incredibly huge, and that sub-sea movement is what generated the tsunami. The tsunami was the most devastating aspect of the quake, and it traveled more than 5,000 miles. It was still 5 feet high in South Africa. The greatest number of casualties was in Indonesia, with about 131,000 deaths, but the tsunami killed at least 35,000 in Sri Lanka, 12,000 in India, 5,000 in Thailand, and nearly 100 on the coast of Africa.

—Richard I. Gibson

Drawing Public domain (US Army Corps of Engineers) ; Satellite image from NASA

December 25. Quaternary Megafauna Extinctions

Mammoths, mastodons, saber-toothed cats, American horses and camels, the giant bear, the Irish elk, Glyptodonts, which were giant relatives of the armadillo, the Megatherium, an elephant-sized ground sloth, giant kangaroos, a 7-foot-long beaver, and dozens of other large animals and animal groups, collectively called megafauna, died off during the Quaternary Period of the Cenozoic Era, many of them at about the time of glacial retreat, 10,000 to 12,000 years ago at the end of the Pleistocene Epoch. Some of them were relatively new to planet earth, and some species had been around for millions of years. 

Don’t visualize this extinction as an instant in time, or really even a short period of a few hundred years. It’s true that there seems to be a spike in extinctions right at the time the glacial period was ending, 10 or 12 thousand years ago, but many animals, including for example the giant terror birds of South America, were becoming extinct at about the beginning of the Pleistocene glaciation, 2.5 million years ago.   

Mammoth (left), American mastodon (right) Image by Dantheman9758, used under Creative Commons license. 

And how do we know something was completely extinct? We can only go by the fossil record, and fossils of modern large animals are rare. Even for a particular species, extinction can take thousands of years, winding down to the last mastodon, or passenger pigeon, or whatever. Mammoths got started about five million years ago, in the Pliocene, and were mostly gone at about the end of the glacial period, 12,000 years ago. But small populations survived on isolated Arctic islands such as St. Paul, off Alaska, and Wrangel, off Siberia, as recently as 3700 years ago. Different species in different niches disappeared sooner or later, depending on circumstances. The pygmy mammoth probably survived on the Channel Islands of California until about 11,000 years ago.

So, why did the mammoths, giant sloths, and all the rest die? The first obvious explanation is the dramatically changing climate. Glacial environments certainly constrain food supplies and habitat, and other habitats, even in temperate zones, change. That could account for some of the extinctions early in the Pleistocene, 2 to 2½ million years ago. But why would so many go extinct at the end of the glaciation? Well, change is change, and warmer might not necessarily always be better for all species; forests supplanted grasslands in the north, perhaps denying mammoths their main food sources. It might be a simple matter of inability to adapt to the changes. Once a population becomes small, many things – disease, a big storm, a flood – could have been the final blow that ended a species that may have been in decline for some time.

The second primary idea for the extinction of the Quaternary megafauna is still somewhat controversial, and involves hunting of the animals by predators – specifically, early humans. There’s no question that humans hunted and ate some of these animals – or hunted them to remove them as threats. Homo erectus was killing mammoths a million and a half years ago. The question really is, were humans the primary, or even a major, cause of the extinctions.

There’s pretty good correlation between the arrival and expansion of humans in various areas with the demise of the large mammals. The correlation in time and space alone does not mean the one caused the other. Arguments against this hunting hypothesis include the idea that the small populations of humans would not likely have been enough to eradicate millions of animals. And while the correlations are best in the Americas and Australia, the extinctions take place at about the same time in Eurasia, where these animals had been exposed to human threats for far longer.

The bottom line is that the ultimate cause of extinction of so many large mammals is uncertain. Humans may have had a role, either large or small, and climate factors are likely to have been important, and perhaps even controlling in the extinctions. When you take multiple climatic factors into account, including temperature changes, precipitation patterns, vegetation and habitat changes, maybe that’s enough to explain most of the extinctions, and humans or disease or something else were simply the exclamation point at the end. In such complex systems as the earth, it’s almost impossible to point to any single thing as THE cause, and as usual, the most likely answer is a combination of many factors. We really don’t know.

—Richard I. Gibson

Giant mammals
Mastodon extinction 
Extinctions affected survivors (e.g. coyote) 
Pleistocene beasts (an excellent blog)
Debunking the impact idea 

Image by Dantheman9758, used under Creative Commons license.   

December 24. Pleistocene Ice Ages

There have been at least five major glacial periods in earth’s history – two in the Precambrian, one at the end of the Ordovician, one during the Carboniferous, and now. “Now” is perhaps an overgenalization, but it’s not all that certain that the most recent ice age is over. It started about 2.6 million years ago, and occupied the Pleistocene Epoch of the Quaternary Period of the Cenozoic Era. It ended – if it has ended – about 12,000 years ago, which is taken to be the end of the Pleistocene and the start of the Holocene, the present-day epoch of geologic time. 

I say we don’t know if it’s over because the 2.6 million years saw at least four major and several minor pulses of glaciation that alternated with times called interglacials, when the ice was much less extensive. We may be in the early stages of an interglacial period, or the whole cycle may be over. 

Maximum glaciation (ice shown in black) by Hannes Grobe/AWI, used under Creative Commons license.   

At the peak, glacial ice reached as far south as the Missouri and Ohio Rivers and New York City in North America, and across the British Isles, central Germany, and most of the plains of western and central Russia. Continental ice sheets as much as 3 to 4 km thick (2 to 2½ miles) were prevalent in the northern hemisphere, which seems to have been affected more than the southern hemisphere, though the Antarctic Ice Sheet expanded. Mountain glaciers in southern South America, New Zealand, and in the northern hemisphere also grew significantly.

Because so much water was locked up in the ice, sea level fell about 120 meters, or 390 feet. Huge areas of what are now the continental shelves, under water, were dry land, including the Bering Sea off Alaska, which created the famous Bering Land Bridge between North America and Asia.

What caused it? Over the course of these podcasts, you’ve heard about ideas of global cooling related to volcanism that might have led to glacial periods, as well as ocean circulation, the rise of plants that changed greenhouse conditions, and other things as possible causes for the sporadic glacial times through earth history. The bottom line is, we really don’t know for certain. The modern glacial period, the past 2½ million years, may be just the culmination of cooling that dates to 34 million years ago, when the Antarctic Ice Sheet got going.

One factor that has been promoted as a cause for ice ages is the variations in earth’s orbit. Regular cycles in the tilt of the earth’s axis and in the shape of the obit around the sun produce regular variations in temperature – but that happens all the time, and there have only been a handful of glacial periods over billions of years, so there must be more to it.

Pleistocene ice extent (USGS)

Ocean currents and the arrangements of continents may have a role. I mentioned the opening of the Drake Passage, creating a truly circumpolar oceanic current around Antarctica, as a possible factor in the growth of ice there. In the modern Northern Hemisphere, the changing arrangements of continents has made the Arctic Ocean an almost landlocked sea, and combined with the growth of the Isthmus of Panama about 3 million years ago that drastically changed global circulation patterns, maybe that all created the tipping point that led to the onset of glaciation.

A decrease in carbon dioxide levels would reduce the greenhouse effect and global average temperatures would be cooler. Such changes can be observed in ice cores, and there was indeed a general, slow, but significant decrease in CO2 in the atmosphere through most of the Cenozoic until recently. That would still call for a threshold or tipping point because the onset of significant continental glaciation was pretty abrupt, about 2.6 million years ago, and it is hard to explain the changing CO2 levels in ways that would account for the rapid changes from glacial to interglacial periods. And even within predominantly glacial or non-glacial times, there was lots of variability. It’s best to think of it as “mostly glacial” periods that lasted on the order of 40,000 to 100,000 years, and shorter interglacial periods, maybe 20,000 to 30,000 years – but with oscillation and change throughout it all.

The last glacial maximum was about 22,000 years ago, and at that time there was more ice in North America than in present-day Antarctica. Most of the estimates of the alternating temperatures and ice volumes suggest a gradual build-up to an ice maximum, over a period of 40 to 60 thousand years, but a pretty abrupt drop into an interglacial, over a period of around 10,000 or 15,000 years. That’s pretty much how long it’s been today since there were glaciers in Ohio.

There are at least 9 professional journals devoted to the study of the Quaternary and the glacial period. I’m going to leave the question of the cause at what I’ve said in this episode, recognizing that there is plenty of research going on. The effects of the ice age are much easier to describe than the causes, and four of our next five episodes will relate to some of those.

—Richard I. Gibson

Impact on Gulf of Mexico
Maximum glaciation by Hannes Grobe/AWI, used under Creative Commons license.    

December 13. Antarctica freezes

First today, for those of you who are into vertebrate fossils, I came across a nice blog report on this year’s meeting of the Society of Vertebrate Paleontologists in Berlin. The blog has a nice summary of new findings, from Archaeopteryx to dinosaur colors, to giant kangaroos. LINK


* * *

Today we’re going to Antarctica. 

Australia and Antarctica began to rift away from each other back in the Cretaceous, probably about 95 million years ago, with the initial faulting and rifting beginning in the Jurassic. The separation was pretty slow for many millions of years, so that by Eocene time the distance between Antarctica and Australia was only about 500 km, or 300 miles. The speed of separation and amount of sea-floor spreading increased in Eocene time from maybe 4 cm per year to 7 cm, and today there’s an ocean basin 2100 miles wide between them, and it’s still growing. 

Within Antarctica, all was not quiet. The West Antarctica Rift was the beginning of a pull-apart between the larger East Antarctic craton and the smaller West Antarctica area. It started probably during the Cretaceous and continues to this day. The flank of the rift is more or less the Trans-Antarctic Mountains today, and the active volcanism there including Mt. Erebus is probably related to ongoing activity along the rift. Climate change together with the uplift of the mountains in central Antarctica led to the development of snow cover there beginning about 34 million years ago. The boundary between the Eocene and Oligocene, 34 million years ago, is marked by a decrease in carbon dioxide levels, and general global cooling, which allowed for the beginning of ice buildup in Antarctica. Before about 34 million years ago, there was no ice in Antarctica even though it was near the south pole.

The last connection between Antarctica and any other continent, South America, also disappeared about 35 to 40 million years ago. With the opening of the Drake Passage between Antarctica and South America, the circumpolar ocean current was established, and its presence prevented significant interchange of tropical and polar waters. This was another factor in the cooling of Antarctica.

The West Antarctic Rift is positioned about where Australia and New Zealand were attached to Antarctica in the old Gondwana. The Tasman Sea, between Australia and New Zealand, began to extend and separate those areas back in the Cretaceous, about the same time as the West Antarctic Rift began. Even though it’s been active for close to 100 million years, the West Antarctic Rift hasn’t really gone anywhere. The active volcanism shows that there is still extension of the crust, breaking, to let the magma up, but an ocean has not formed between West and East Antarctica. The Tasman Sea spread for about 22 million years, 80 to 58 million years ago, Cretaceous into the Paleocene epoch of the Cenozoic, and then apparently stopped. Today, Australia, the Tasman Sea, and about half of New Zealand – the part west of the Alpine Fault – are together on the same tectonic plate. The whole package is pushing against the Pacific Ocean Plate, making for some complex, violent interactions along oceanic trenches north of New Zealand and in collision zones in New Guinea. The oceanic part of the Australian Plate west of Australia itself is subducting beneath the southern part of Asia, in Indonesia.

So – the bottom line for Antarctica is that two things – tectonics and the positions of continents, together with climate change and general cooling, led to the beginning of the Antarctic ice sheet about 34 million years ago, about at the boundary between the Eocene and Oligocene epochs of the Cenozoic. It’s had a long time to build up its ice – much longer than the ice ages that happened in the past 2 million years or so.

—Richard I. Gibson

Australia-Antarctic separation 

West Antarctic Rift System
Trans-Antarctic Mountains 

Map from NASA, annotated by Gibson (generalized, based on various sources). 

August 17. Glaciers and Coal

Today’s episode is a response to a listener’s question about the close juxtaposition of glacial deposits and coal beds in Australia.

Despite the abundance of coal in the Carboniferous, especially in the northern hemisphere, and despite the changing climate that meant coal formation there largely ended with the end of the Carboniferous, there’s plenty of Permian coal too. Most of it is in the former Gondwana – Australia, South Africa, India, South America, and Antarctica, but there is a lot of Permian coal in Russia as well. In Gondwana, the coal is pretty closely associated with glacial deposits.

We might expect that glacial deposits and coal swamps would reflect two very different environments, but so far as I can tell, they pretty much co-existed in Permian time at least in quite a few places.

In Australia and elsewhere, the coal-bearing rocks and glacially deposited layers actually interfinger. So at best, we might have had some relatively rapid changes in climate to switch from glacial times to warmer, coal-swamp times, and from what I read there were at least 8 specific glacial periods in the late Carboniferous and early Permian.

But the alternative explanation, and from what I can gather it seems to be the preferred one, is that these areas were on the margin of the ice, and plant life actually thrived there. The keys to making coal are 1) lots of plants and 2) rapid burial of the plant matter so it does not have time to decay. Our typical vision of warm swampy areas with low oxygen to prevent decay is just one way to do that. A cold climate, with plants buried by glacial debris, would work just as well, if not better.

Gondwana base map from Du Toit (1937, Our Wandering Continents). Blue line is highly generalized margin of glacial area; solid black are highly generalized coal deposits (based on Langford, 1992).

I think that while Australia was certainly part of the glaciation in the southern polar part of Pangaea, it was probably far enough away from the pole (which was more or less in South Africa, but there are coal deposits there, too) that the climate might not have been like modern Antarctica, but perhaps more like modern Patagonia but with glaciers. So abundant plant life could have been growing, even thriving, near the glacial margin. A modern analogy would be the peat bogs of temperate and even arctic climates. When glaciers receded, forests and peat on the tundra would advance. When glaciers advanced, the deposits the glaciers carried would have buried the forests. This would be a good way to get the interfingering of glacial and coal deposits that we do observe.

This would not strictly be cyclothems, which represent rises and falls of sea level, alternately allowing swamps to form and then burying them in river sediment, but they would be cyclic nonetheless, like cyclothems. I do not know if the coal-glacial sediment packages follow the 8 known glacial periods or are (likely) something more complex, but if they do it would be on a periodicity of a few million years. Standard cyclothems can show alternations that may represent changes on scales of a few tens of thousands of years, or even fewer, as well as the longer periods of millions of years. If the coal results from glacier-margin plant life, as I infer it does, then the alternations would not reflect sea-level changes as cyclothems do, but more directly would reflect changes in position of the glaciers (together with the dumping of sediment to bury the forests or other vegetation).

—Richard I. Gibson

Links:
Permian of Australia
Permian coal in South Africa 

Langford, 1992 - Gondwana’s Permian coal 

August 9. Permo-Carboniferous glaciation

Glacial development in the southern continent of Gondwana began during the Carboniferous. We attributed the cyclic nature of coal cyclothems to alternating high- and low-stands of the sea related to advance and retreat of glaciers. The glacial period lasted into the Permian, so it is usually called the Permo-Carboniferous glaciation, and it lasted close to 90 million years, the longest glacial epoch in Phanerozoic history, the past 550 million years.   

Some of the best evidence for the glaciers comes from South Africa, where rocks called tillites are many hundreds of feet thick. Till is the poorly sorted deposit of material left behind by a glacier, and is has a distinctive chaotic character. You can get till-like deposits in various ways, however, but in addition the cobbles and pebbles within till are often striated – scratched by other rocks held firmly in glacial ice. Together with broader geometry of the deposits, including striated pavements over which glaciers carrying rocks flowed, we’re sure that these sediments were laid down by glaciers.

Permian Gondwana reconstruction and inferred ice cap (blue outline) from A. du Toit, South African geologist, 1937 (Our Wandering Continents). Du Toit's work was remarkably prescient. The map above differs only slightly from modern reconstructions of Gondwana.

The tillites in South Africa have been well known for more than 100 years. South Africa lay in the heart of Gondwana, well inland and probably at a relatively high elevation, and not far from the South Pole. This made for conditions favoring snow and ice accumulation. The ultimate causes of the Permo-Carboniferous glaciation aren’t completely clear, but the presence of a large polar continent is almost certainly an important factor. Other factors include the overall geometry of seaways and precipitation patterns. As Pangaea formed and equatorial seaways closed, oceanic circulation changed. Interiors of continents, distant from the sea and especially in mid-latitudes like the Sahara today would see low precipitation, and in cold climates the precipitation that did fall would likely be snow.

Details of stratigraphy in the South African tillites suggest that there may have been at least four distinct glacial periods within the overall glacial epoch, lasting 5 to 7 million years each, with intervening interglacial periods.

Water frozen in ice obviously affects sea levels, as we’ve said, but it can also affect carbon dioxide levels in the atmosphere by locking oxygen in water into the ice. Or was it the other way around? In late Carboniferous time CO2 levels were falling dramatically, perhaps partly as a result of the vast extent of plant life around the globe. Did that reduce greenhouse conditions, cooling the planet and enhancing the growth of glaciers? The glacial maximum appears to have been in early Permian time, and the glaciation seems to have pretty much ended later in the Permian, coinciding with a rebound in CO2 levels. The interplay of all these factors is still under considerable investigation.

Rocks virtually identical to the glacial deposits of South Africa were found in South America, East Africa, Antarctica, India, and Australia, and this discovery was another nail in the coffin helping to define the supercontinent of Gondwana. Together with fossil evidence like glossopteris that we talked about yesterday, the concept of a continuous sheet of ice was important to the idea that the southern continents had once been assembled into one landmass.

* * *

On this date, August 9, 1138 a.d., an earthquake destroyed the city of Aleppo, Syria. The generally accepted death toll, 230,000, makes it one of the deadliest earthquakes in history. Aleppo sits more or less at the end of a transform fault that is the boundary between the Arabian Plate and Africa, right where that fault impinges on the Anatolian Plate in Turkey, a small block that is more or less amalgamated to the Eurasian Plate. This tectonic activity is obviously modern, but its heritage goes back to the Permian, when Africa, as part of Gondwana, was beginning to impact Eurasia. It’s still doing that, 270 million years later.

—Richard I. Gibson

Links:
Permian glaciation 
Permo-Carboniferous glaciation
Gondwana glaciation 
Crowell, 1978

June 25. Glaciers in Australia and South America

After the end of the glacial period at about the end of Devonian time, most of the Mississippian was ice-free around the world until near the end of the period. There is evidence for episodic glaciation in Australia and South America at this time.  

It’s possible that the Mississippian glaciers were predecessors to the well-defined Permian glaciation, although that ice age was still at least 25 million years in the future as the Mississippian ended. The onset of glacial episodes would have contributed to sea-level changes that in turn would be contributing factors in the increase in clastic sediments in Late Mississippian time. The shallow seas so common earlier in the period would have become reduced in size as sea levels fell.

One possible contributing factor in the beginning of the glacial period is the tectonic closing of the Tethys Sea and the ocean between Gondwana and North America. Disrupting what was a major equatorial ocean circulation pattern could have impacted global distribution of heat in the oceans and therefore in the atmosphere, and could have helped continental ice sheets to form and grow.

—Richard I. Gibson

Technical paper

March 28. Late Ordovician Glaciation

I’ve mentioned the late Ordovician glaciation several times in recent posts, suggesting things that might have been factors causing it. One possible factor is the fact that the largest continent, Gondwana, which included most of Africa, South America, India, Australia, and Antarctica, was located over the south pole. That alone wouldn’t do it unless the climate was cold enough.

Upsala Glacier, Argentina

During most of the Ordovician, planet Earth was in greenhouse conditions, with high concentrations of carbon dioxide in the atmosphere. We mentioned two things that might have affected that – the presence of life on land, taking more and more CO2 out of the atmosphere, and the tectonic activity that made a high mountain range in what is now eastern North America. Erosion of that mountain range, to create the immense Queenston Delta, might have produced enough sediment to have an effect on the atmosphere so that CO2 was reduced. Both of those ideas are in the “might have” category, but something certainly did affect CO2 concentrations.

And then we talked about the humongous volcanic eruptions, the Deicke and Millbrig and others near the beginning of the Late Ordovician epoch, around 455 to 457 million years ago. Cooling because of high concentrations of dust in the atmosphere was certainly likely, and the timing is good with respect to the glaciation.

The glaciation seems to have been at its height about 440-455 million years ago, which includes the first part of the Silurian – the boundary with the Ordovician is put at 443 million years ago. There’s some evidence that the glaciation might have started as long ago as 460 million years.

The evidence about CO2 values isn’t speculative. There is abundant evidence in carbon isotopes to indicate significant changes in water temperature – a pretty significant disruption of the carbon cycle, enough to imply cooling and a reduction of the greenhouse effect. In addition to the ideas that life and erosion helped remove CO2 from the atmosphere, another idea is that volcanism helped. But wait, you say – I thought volcanoes added CO2 to the atmosphere. Well, they do. But earlier in the Ordovician, extensive basaltic eruptions – when they were finished – created large expanses of solid basalt, which reacts and erodes relatively quickly, and might have been a factor in CO2 removal. That’s another pretty speculative possibility.

Glacially-deposited rocks of Late Ordovician age are common across what is now the Sahara Desert, from Morocco to Ghana to Libya, and they’ve also been found in South Africa, Brazil, Arabia, Germany, Nova Scotia, and Newfoundland. The ages of all those deposits are not absolutely determined, but they are all of Late Ordovician or Early Silurian age, coinciding nicely with and defining the glacial period.

There’s evidence that this glacial event ended quite abruptly, for reasons that seem to me highly speculative and hardly able to explain it. One suggestion is that once the ice sheets reached their limit, they essentially collapsed – which to me doesn’t adequately explain why they then retreated to practically nothing. CO2 levels increased in the Silurian, but I haven’t seen a good explanation for why.

The Ordovician-Silurian glaciation is the only one associated with a mass extinction event – one of the largest in earth’s history. We’ll talk more about that at the end of the Ordovician, in a few days.

—Richard I. Gibson

Further reading:
The carbon cycle
Photo by longhorndave via Wikipedia, under cc-by-2.0