Mammoths in the insular Nearctic? Some constraints on the existence of a Pleistocene megafaunal refugium in the Canadian Arctic Archipelago

• Published in: Quaternary international 2007-07, Vol.169, p.29-38
• Main Author: MacPhee, R.D.E.
• Format: Article
• Language: English
• ISSN: 1040-6182; 1873-4553
• DOI: 10.1016/j.quaint.2006.08.007

Mammoth fossils recovered on Banks and Melville Islands in the western part of the Canadian Arctic Archipelago have been used as a basis for inferring that these islands functioned as a Late Pleistocene refugium. However, only two fossils have actually been found in these locations, and both date to the period immediately before the Last Glacial Maximum (LGM). This paper presents a series of paleontological, chronometric, climatic, biogeographical, and glaciological constraints on scenarios that seek to explain how mammoths (or their fossils) managed to reach the western archipelago. In combination, the constraint set strongly contradicts the recent argument that living elephants dispersed to Banks and Melville by walking across the exposed seafloors of Amundsen Gulf and M’Clure Strait. Swimming cannot be discounted, although in such a scenario colonization would have most likely taken place during a warm period (Sangamonian or one of the major Wisconsinan interstadials), when interisland channels would have presumably been ice-free during the summer. But in that case elephants should have been present on the islands well before the LGM—and consequently there should be more fossil evidence establishing their presence. Alternatively, the mammoth fossils may have been deposited by ice-rafting, long after the deaths of the animals themselves. But in that case there should be many more examples of ice-rafted megafaunal fossils on Arctic islands. Ultimately, the scantiness of the evidence precludes making a decision about how or whether living mammoths reached the islands; the only solution lies in more extensive collecting in the High Arctic.