Extinction of the mammoth and the Clathrate Gun

Print Friendly, PDF & Email

Joanne Ballard and André Bijkerk (USA)

In this article, we will argue that the extinction of megafauna on the mammoth steppes of the Northern Hemisphere may ultimately have been caused by the release of massive quantities of methane in the North Atlantic Ocean at the Amazon Fan near the Brazilian coast and also from the Ormen Lange gas field off the coast of Norway. We will suggest that these events caused significant changes in the flow of water at the surface of the ocean that, in turn, led to very rapid changes in the levels of rainfall.

Scientists have already recognized that increased precipitation gave rise to changes to ecosystems (or, more precisely, to biotopes) that destroyed the mammoth steppe. However, much of the evidence we will use in this article to support our argument has been used to support other sorts of explanation for the extinction. Therefore, this primary evidence now appears to be in need of revision.


About 11,000 years ago, all of the remaining herds of mammoths suddenly disappeared. During the Pleistocene, these mammoths once thrived on a vast, megafauna steppe stretching across the Northern Hemisphere. It may have resembled the African steppes of today with lions, hyenas and several species of large grazers being present. However, the debate about the cause(s) of the extinction continues. In North America, things appear to be simple – the appearance of early humans on that continent seems to coincide with the downfall of the megafauna. However, there are also profound discrepancies in the details.

In Siberia, where the woolly mammoth was one of the predominant species, very few humans were around, yet the extinction was just as rapid as in North America. Moreover, there are clear signs of rapid and drastic climate changes in Siberia. A sudden shift in the climate changed the dry, treeless, mammoth steppe into marshes, swamps and forests probably covered with thick snow in winter – a habitat completely unsuitable for mammoths. However, the big question is, “What caused these sudden habitat changes?”

Fig. 1. Synthetic hydrate, formed in the laboratory of Gerogia Tech Professor J. Carlos Santamarina, sustains a flame. (Photo courtesy Dr Carlos Santamarina.)

Ice core observations

To answer this question, a thorough review of the literature on paleoclimate is required, particularly articles concerning the time around the end of the Pleistocene. This takes us first to the ice cores of Greenland, where we see rapid changes. Figure 1 shows how the heavy oxygen isotope ratios (δ18O) of the snow appear to follow a roller coaster ride. (These isotope ratios are referred to as “proxies” that are measurements that can be used to represent something else, in this case, paleoclimate change.) We also see that, when the isotopes go up at the beginning of the Bølling Allerød interstadial (about 14,500 years ago) as well as the Preboreal, which is the start of the current epoch referred to as the Holocene (about 11,700 years ago), the snow accumulation and the methane concentration do as well.

Fig. 2. Comparing the timing of the d180, relative layer thickness and methane concentrations of Greenland ice cores.

These graphs have become rather familiar in climate change studies and the prevailing viewpoint is that rapid changes in global temperature caused the changes in rainfall. In turn, this caused swamps to proliferate, which then increased the methane in the atmosphere. However, there are some problems with this view. We know that the Siberian mammoths appear to have thrived in the alleged cold areas existing at the times of the Last Glacial Maximum (about 18,000 to 22,000 years ago) and during the Younger Dryas (about 12,750 to 11,7oo years ago).

And, of course, woolly mammoths do not have that name for nothing! They must surely have been capable of enduring extremely cold winters. Nevertheless, the grass of the steppes needs a certain minimum temperature to grow and to provide 150 to 300 kg of fodder per mammoth each day, all year round (on the assumption that their metabolism was comparable to closely-related modern elephants). This fact alone assures us that the summers during these periods were not that cold.

Oceanic observations

Ice cores are only one of many sources of paleoclimate information. Since oceans, which are also the main source of that increased precipitation, cover most of the Earth, we certainly need to investigate oceanographic evidence as well. Sure enough, the analysis of ocean sediments also shows considerable and rapid changes, with the peaks and valleys of the graphed oceanic data shadowing those of the ice cores.

It is clear that large changes have taken place in the oceanic currents as well, resulting in changing sea surface temperatures in certain places. We see that this had consequences on a much larger scale. For example, when the oceans started to act up at the beginning of the Bølling Allerød (about 14,500 to 12,750 years ago), Africa entered into a very humid period that lasted several thousand years. North America, where large ice sheets were melting, also shows evidence of distinct, dry and wet periods.

Fig. 3. Methane hydrate on the sea floor as home for worms. (Photo courtesy of Ian R. MacDonald, Texas A&M Univ. Corpus Christi, Texas, USA.(

Apparently, these oceanic changes had a large impact on rainfall patterns, especially in the Northern Hemisphere. We see similar changes in precipitation connected with the well-known phenomenon of the warm, cyclical El Niño current in the Pacific along the equator. But what causes the oceans to show such unusual, massive and abrupt variations? Given the enormous inertia of the oceans, which react slowly to external changes, it appears evident that these abrupt events are difficult to explain by external causes. Rather, it appears that something happened directly to the oceans that had a secondary effect on climate. Unusual events may have just as unusual causes and, in this instance, we might take a look at the large methane hydrate fields on the outskirts of the continental shelves.

Methane hydrate

Methane hydrate has a complicated, cage-like shape, with methane gas inside a structure of frozen water molecules. This results in the strange phenomenon of ice that is flammable. It can only exist at low temperatures and high pressures, but can be found on the ocean floors in large quantities. However, when the water temperature rises or the pressure of the water column drops sufficiently, the hydrates become unstable and the methane forms gas bubbles that rise to the ocean’s surface. The methane on the ocean floors is formed mostly by the decay of organic matter but can also originate from natural methane gas seepage from the earth’s crust. The latter form can be gaseous so long as the higher geothermal temperatures prevail but may be converted into hydrate by contact with very cold water at the bottom of the ocean.

Fig. 4. The bubble stream in a aquarium entrains water with it, circulating the water in the tank. Consider the effect of the bubbles from a huge methane hydrate release from the ocean floor entreating cold bottom water and bringing it to the surface.

This is all very well but how could methane hydrate have killed the mammoths?

Suppose that a very large hydrate field starts to decompose and a dense bubble stream starts to rise to the surface, forcing the water up with it. Such an event has been proposed before and is called the “Clathrate Gun hypothesis”. This same scenario can be observed in an aquarium. The bubble stone creates a bubble stream that causes the water to flow throughout the whole fish tank.

The upwelling water forces the surface waters to flow away. If something similar were to happen at the equatorial area of the North Atlantic, it would send the warm tropical surface waters northward, which could have a definite impact on the climate for the duration of the event. If a bubble stream event (hydrate release) were to happen further north of the equator, the relatively warmer Atlantic surface waters would penetrate deep into the frigid Arctic Ocean, also causing dramatic changes in rainfall patterns.

The evidence

How can one prove all this? One must look at the data collected by hard-working scientists doing research in these areas to see if there is anything that refutes the hypothesis.

The best evidence of these possible methane release events would be to find the sources and see if they match with such a scenario. Sure enough, there are huge, concentrated methane hydrate fields in the ocean beyond the mouth of the Amazon River, the alluvial fans where the remains of the Amazon rainforest accumulate and decompose. Sedimentation rates amount to about 1.5cm a year, enough to build the potential energy of large hydrate bombs in a few ten thousand years. T

he last big eruption of hydrate here was some 13,000 to 14,000 years ago, perfectly matching the methane spikes and precipitation maximums in the Greenland ice cores during the Bølling and Allerød interstadials. When these outbursts ended, the spikes in virtually all of the previously mentioned ‘proxies’ (the ice cores and the ocean sediments) dropped again. The result was to leave the ocean surface with very low surface temperatures after the cold water at the bottom of the ocean had risen to the surface, especially in the Northern Hemisphere. This disrupted the normal water evaporation cycle and it became very dry on a large scale, a period known as the “cold Younger Dryas”. It would appear that these sudden climate changes were responsible for the extinction events both in Europe and North America but not in Siberia, which was apparently too far away from the Amazon events.

Fig. 5. Rachel Brown, geoscience student from Indiana University Southeast in the bonebed at the Mammoth Site Museum in Hot Springs, South Dakota.

Then a second, much larger event occurred. A large methane hydrate field off the coast of Norway started to erupt some 11,000 to 12,000 years ago. Considering the very active methane seepage in this area, it is likely that this methane was mostly of natural gas origin rather than from decomposing biota. This eruption coincided with the beginning of the precipitation changes of the Preboreal at the end of the Younger Dryas.

The now-atmospheric methane, along with snow precipitation, was deposited on the Greenland ice sheet and recorded in the ice cores. In addition, this methane hydrate event forced the relatively warm Atlantic waters into the Arctic and the ensuing precipitation signalled the end of the remaining mammoth steppe in Eurasia along with many of its inhabitants. The eruption cycles ended some 8,200 years ago. Evidence of this consists of a huge scar of submarine sediment displacement – the largest turbidite on record. This is known as the Storegga slide that is believed to have caused a tsunami reaching a height above sea level in excess of 20m on the Shetland Islands.


Although the body of available evidence fits very well into such a hypothesis, there is a slight complication. Almost all of the data has already been used to support other sorts of explanation of the last glacial transition. Therefore, we now face the task of having to unexplain these explanations. For example, the isotope spikes in the ice cores are thought to represent warm-cold transitions instead of dramatic precipitation changes as we suggest.

Fortunately, it has been recognized for a long time that these isotope changes could also be caused by changes in rainfall over the course of the different seasons in a year. However, this idea has been previously rejected on the basis of climate modeling, as a result of the apparent absence of evidence for all the complications that this other explanation gives rise to. However, more and more articles are now emerging that report much earlier warming than the Greenland ice cores reveal, and many articles about the Younger Dryas appear to confirm its dry character much more than its cold character. It could have been cold in many places but not as cold as the ice cores suggest. Moreover, several studies report warm summers during that period.

Digging further, it appears that we do indeed encounter a controversy about the nature of these isotope spikes. It is becoming increasingly clear that the Northern Hemisphere warming, which occurred after the Last Glacial Maximum, was much earlier (about 17,000 years ago) than the Bølling Allerød spikes in the ice cores had suggested (14,500 years ago) and happened at the same time as the post-glacial warming of the Southern Hemisphere. However, almost identical isotope spikes are recognized in several other sediment proxies over the Northern Hemisphere and, therefore, are too late to represent that warming. So, it becomes apparent that some re-considerations of the data are in order.

A factor that has not been readily recognized before is the magnitude of the effect of methane hydrate decomposition events on climate. The research has focused on the release of methane as a greenhouse gas and on isotope signatures. Several studies have concluded that methane hydrate events could have had only a very marginal effect. However, we have found no previous discussion of the physical effects of methane hydrate on the ocean currents and the secondary effects on many other processes (for instance, the so-called “meltwater pulses”).

Another unrecognized effect is the upwelling water in the methane bubble stream coming from the seafloor having much higher carbon dioxide concentrations. As this bottom water surfaces, pressure diminishes and the carbon dioxide is released in a similar way to when the cap is taken off a Coke bottle. Indeed, the clearly visible carbon dioxide spikes in the Antarctic ice cores could have been caused in just this way.

Fig. 6. The woolly rhino (Coelodonta antiquitatis), once thriving all across Eurasia on the mammoth steppes, disappeared suddenly during the Belling Allered interstadial. (Replica sculpture by Remie Bakker, Manimalworks Rotterdam NL, Picture courtesy of Rene Bleuanus, SQZI Concept Studio, Arkel NL).


In conclusion, we suggest that the mammoths and other megafauna became extinct as a result of sudden increases in rainfall that ruined the vast, dry steppe. The precipitation changes were brought about by sudden, strong changes in the flow of surface seawater. In turn, these were caused by massive, bubbling water–methane gas chimneys from the decomposition of the methane hydrate fields of the Amazon Fan and the Norwegian submerged gas fields in the Ormen Lange region. However, reinterpreting the abundant evidence that currently exists from “sudden abrupt warm–cold variations” to “sudden abrupt humid–dry variations” will be a substantial task and the resulting debate promises to be somewhat ‘robust’! However, this hypothesis does have the advantage of explaining the most phenomena and giving rise to the least contradictions. For these reasons, it may, therefore, prevail.

And, none of the other explanations of events of the ice ages gives a second thought to the extinction of the mammoth, the Pleistocene ‘poster child’ of catastrophic climate change!

Non calor, sed umor, it’s not the heat, it’s the humidity.

Joanne Ballard is a Geosciences Senior at Indiana University Southeast and André Bijkerk is an independent Researcher.

Fig. 7. Joanne Ballard and composite skeleton of woolly mammoth, constructed by Dick Mol and others, at Naturalis in Leiden, The Netherlands.


The authors want to thank Dick Mol, Mammuthus International, the Netherlands; Dr. Larry Agenbroad of the Mammoth Site, Hot Springs, South Dakota, USA; Dr. Norm Catto, University of Newfoundland, Canada; and Dr. Glenn Mason, Indiana University Southeast, Indiana, USA for their support and encouragement to follow this line of inquiry into the extinction of the mammoths.

Selective references

Guthrie R.D., 2001. Origin and causes of the mammoth steppe: a story of cloud cover, woolly mammoth tooth pits, buckles, and inside-out Beringia. Quaternary Science Reviews 20, pp. 549–574.

Hubberten et al., 2004. The periglacial climate and environment in northern Eurasia during the Last Glaciation, Quaternary Science Reviews 23, 1333–1357; Kennett J.P. et al., 2003.

Methane hydrates in Quaternary climate change: The clathrate gun hypothesis. Publisher, American Geophysical Union, Washington, DC; Maslin, M.A., C. Vilela, N. Mikkelsen and P. Grootes, 2005.

“Causation of the Quaternary catastrophic failures of the Amazon Fan deduced from stratigraphy and benthic foraminiferal assemblages.” Quaternary Science Review, volume 24, Issue 20-21, 2180-2193; Mol D. et al., 2006.

Results of the CERPOLEX/Mammuthus Expeditions on the Taimyr Peninsula, Arctic Siberia. Russian Federation Quaternary International, January volumes 142-143 pp. 186-202; Mienert, J., Vanneste M., Bunz S., Andreassen K., Haflidason H., and Sejrup H.P., 2005.

Ocean warming and gas hydrate stability on the mid-Norwegian margin at the Storegga Slide. Marine and Petroleum Geology, 22 (1-2), pp. 233-244; Nisbet E.G, 2002.

Have sudden large releases of methane from geological reservoirs occurred since the Last Glacial Maximum, and could such releases occur again? Philosophical Transactions Royal Society, London A 360, 581(607).

Leave a Reply

%d bloggers like this: