During the Last Glacial Maximum, Switzerland was almost entirely covered with ice. However, in the ice-free region situated less than 50 km north of the glaciers, human occupation is confirmed as early as 23000 cal BP. Numerous sedimentary sequences in lakes and mires have produced a wealth of proxy data indicating that environmental conditions improved rapidly after the melting of the glaciers that liberated the Swiss Plateau at least at c. 17500 cal BP, offering severe but possible life conditions to plant, animal and human communities. Contrary to what has long been the prevailing opinion, Magdalenian re-colonization of Switzerland did not start with the onset of the warming of Greenland Interstadial 1e, but well before. According to most of the recently obtained AMS-dates, the Magdalenian occupation falls within the cold, treeless, environment of the Oldest Dryas period; it is even conceivable that it did not extend into Greenland Interstadial 1e. More than 50 sites, among which famous caves and rockshelters such as Kesslerloch and Schweizersbild, as well as large open-air campsites like Monruz and Moosbühl, have produced different techno-assemblages that find good comparisons in the rest of Europe. In contrast to the exploitation of mainly local and regional flint sources, the use of "exotic" ornamental/symbolic objects -fossil mollusks, amber and jet -shows widespread, multidirectional long-distance connections with the upper Danube basin, the Mainz basin, the Paris Basin, the Atlantic coast, the Mediterranean and even the Baltic regions.

The two Magdalenian open-air sites Champréveyres and Monruz, on the northern shore of Lake Neuchâtel, were occupied during the warm season, predominantly during the spring and early summer. The diverse faunal assemblages include the horse as the dominant species, which yielded 70 and 90 % of the weight of acquired animals, respectively. Due to their behavioural adaptation as animals with long fleeing distances, horses cannot be hunted close to an actively occupied site. Horse hunts must therefore have taken place at a relatively long distance from base camps and, in addition, could be carried out only from some kind of ambush or hunting hide. Since the weight of a horse precludes transport as a whole, and the skeletal representation indicates that entire horses were present on the site, the most plausible conclusion is that camps were relocated to a successful horse-hunting site. While consuming the horses and possibly preparing them for later use, other activities were performed, including the hunting of other, smaller species. After a mean duration of from one to a few weeks, the camp was relocated to a subsequent horse-hunting site, at a distance of some kilometres. The number of burning cycles in the hearths as well as the amount and types of lithic material corroborate these conclusions. Since horse-hunting sites need to meet specific, strategic characteristics, the number of reoccupations of the main horizons (6 for Champréveyres, 20 for Monruz) is easily explicable.

No abstract