THE BELA PROJECT: LANDSCAPE OF THE BERING LAND BRIDGE AT THE LAST GLACIAL MAXIMUM

After many decades of scientific investigation, the Bering Land Bridge still fascinates Quaternary paleoecologists. In particular, its nature during the last glacial maximum (LGM) remains elusive, as much of the land bridge now lies under water. During the 1990’s, we investigated a paleo-landscape buried ca. 21,000 years ago by a tephra associated with a maar eruption on northern Seward Peninsula. The buried surface, which was incorporated into the permafrost, provided materials for a multi-disciplinary study of soil, surface microtopography, pollen, in situ vegetation represented by plant macrofossils, and the remains of vertebrates and invertebrates. Plant and animal remains and permafrost features all indicate a tundra landscape. The surface supported species-rich, herb-dominated vegetation (with predominantly tundra affinities), and a dense ground layer of acrocarpous mosses; prostrate willow mats occurred in the lee of small topographic features. Pedogenesis was minimal as the surface was subject to a continuous influx of loess, which may have enhanced the nutritive quality of the vegetation and suppressed peat-forming, pleurocarpous mosses. Active layers were similar to modern (ca 50 cm). Chemical and morphological soil features suggest seasonal alternation of dry and moist conditions; this inference is supported by a mixed insect fauna of both xeric- and mesic-adapted taxa. The buried surface of gently rolling interfluves and shallow valleys probably typifies zonal conditions; pollen spectra derived from the 21 ka-surface strongly resemble coeval assemblages derived from lake sediments on both sides of the land bridge, suggesting the latter reflect regional vegetation similar to that of the buried BELA landscape. The study provides the most detailed information to date on LGM land-surface characteristics for this unglaciated high-latitude region, and the results suggest that dry tundra over a relatively shallow active layer is the most appropriate zonal land cover for ecosystem and/or climatic modeling of the LGM.