FOUCAULT, A. and NOBLET-DUCOUDRÉ, N. de (2005). Climate change at the origin of Eurasian mammoth extinction? 2nd International Congress "The World of Elephants". Hot Springs (SD), USA. 22-25 sept 2005. Short papers and abstracts, pp. 55-56.

Climate change at the origin of Eurasian mammoth extinction ?

Alain Foucault, Muséum National d'Histoire Naturelle, Dépt. Histoire de la Terre, 43, rue Buffon 75005 Paris, France, foucault@mnhn.fr
Nathalie de Noblet-Ducoudré, Laboratoire des Sciences du Climat et de l'Environnement, Bat. 709, Orme des Merisiers, 91191 Gif-sur-Yvette, France, noblet@lsce.saclay.cea.fr.

    About twelve thousand years ago, Mammuthus primigenius, the wooly mammoth, vanished from Earth. In Eurasia, some other big species ceased to exist at the same time, among them Coelodonta antiquitatis, Bison priscus, Megaloceros giganteus, Ursus spelaeus, Panthera leo spelaea, Crocuta spelaea. Other big species changed their habitats, for instance Ovibos moschatus and Saiga tatarica.
    Several hypothesis have been proposed to explain these extinctions. Three were recently discussed. Without an order of preference, the first one is that hunting has been fatal for big species, the second one is that a lethal virus killed the mammoths, the third one is that climate changes put an end to the big mammal fauna, not directly, but through substantial modifications of their environment. We evaluate here the latest hypothesis for Eurasia. Environmental changes that occurred after the Last Glacial Maximum (LGM) could indeed be a good reason to explain these extinctions.
    Two complementary approaches can be used for this evaluation. The first one focuses on the reconstructed paleo-environnement from discrete analysis of sediment records, essentially the palynologic content, and on geomorphic arguments. The drawback of this approach is the lack of regional distribution of such data, and therefore the impossibility to reconstruct coherent changes over large geographical regions. The second one uses climate and biosphere models. Climate and biosphere models use variations in insolation as input data, and simulate all atmospheric variables, e.g. temperature, precipitation, as well as the distribution of vegetation that is in equilibrium with the simulated climate. The main problem scientists are faced with, using this approach, is the potential lack of important feedback in the modeling framework that has been built. This is why many studies are devoted to comparing model results to data at different time periods, to assert the realism of the model used.
    Maps issued from modeling are far more accurate in space and time than the ones issued from sedimentological records but some uncertainties remain about the repartition of biomes.
    In our case, if we compare the maps resulting from modeling presented here (see figure) with maps issued from palynologic and geomorphic data elaborated by other authors (1), we can see in some areas, as in western Europe, that, during the LGM, 21 000 years ago, the first ones showed forest whereas the second ones indicated steppes. These discrepancies could be attributed either to inaccuracies in mapping or to differences in the definitions of biomes. Anyway, both methods of mapping lead to the same general conclusions.
    At the time of the LGM, all the non-glaciated parts of Eurasia were covered by tundra, tundra-steppe or steppe. This steppe is often called “mammoth steppe” (2) because of its specific composition that seems to have been particularly appreciated by this big mammal. Many authors have underlined that such a large extent of this herbaceous association was the major reason for the great abundance and variety of herbivores that roamed this area at this time. Abundance and variety of carnivores ensue from the presence of their prey.
    The climate change that occurred after the LGM was characterized by milder and wetter conditions less favorable for steppe growth but much more favorable for forest expansion. Consequently, forests that had taken refuge in few areas during the LGM, gradually spread out at the expense of the steppe. All maps of vegetation distribution show that, 9000 years ago, there were very few remaining areas covered with steppe vegetation in Northern Eurasia. The Northern part was occupied by a tundra, bordered to the South by a large belt of forests, deciduous forest to the West, and boreal forest to the East.
    Without the steppe, or at least without larges areas of steppe in the Northern part of Eurasia, the former association of herbivores could not continue to exist and we are allowed to assume that these environmental changes have been sufficient to explain the extinction of several big species and the emigration of others. However, it is difficult, in the limited precision of our models, to certify that areas of steppe did not remain here and there, and could have been refuges for big herbivores as it was the case for St Paul (3) and Wrangel (4) Islands where mammoths survived until 8000 and 3700 years BP respectively. For this reason, we cannot reject the hypothesis that hunting was the ultimate factor that put an end to mammoth’s life. But in this case, the same explanation has to be true for the other species that disappeared with them.

    In conclusion, climate and environmental changes that occurred after the LGM seem a sufficient circumstance to explain the extinction of mammoths and associated faunas. Modeling of biome distribution must be improved to test the hypothesis that little populations of these mammals could have remained in refuges and eventually could have been killed by hunting.

(1) e.g.Petit-Maire, N. and Bouysse, P. (coord.), 1999. Maps of the World Environment during the last two Climatic Extremes. Commission for the Geological Map of the World and ANDRA.
(2) Guthrie, R. Dale, 1990. Frozen Fauna of the Mammoth Steppe: The Story of Blue Babe. The University of Chicago Press.
(3) Guthrie, R. Dale, 2004. Nature 429:746-747.
(4) Vartanyan, S. L., Arslanov, Kh. A., Tertychnaya, T. V., Chernov, S. B., 1995. Radiocarbon 37 (1):1-6.

Figure. Biome distribution 21 ky and 9 ky ago.