CLIMATE AND VEGETATION IN NORTHEASTERN ALASKA 18,000 YR BP - PRESENT

Changing abundances of taxa in the pollen record of northeastern Alask a contain a climate signal and may be compared with GCM simulations of paleoclimates. Cold, dry conditions indicated by full-glacial pollen spectra are in broad agreement with model simulations. Successive expa nsion of Betula and Salix, and later, Populus, during the deglacial pe riod (14,000-11,000 yr B.P.) are likely correlated with the crossing o f temperature thresholds determined largely by the size and position o f the Laurentide ice. A degree of asynchroneity between sites probably reflects differential site responses to regional effects plus inaccur acies of radiocarbon dating. The early Holocene was probably the warme st period of the present interglacial, as summer climate was influence d by a strong positive insolation anomaly. Picea glauca expanded in ma jor river valleys ca. 9000 yr ago and slightly later at upland sites i n the Brooks and Alaska ranges. At some sites a Juniperus subzone, whi ch suggests particularly warm and dry conditions, occurs between ca. 1 0,000 and 8000 yr B.P. Early mid-Holocene expansions of Alnus (8500-70 00 yr B.P.) and Picea mariana (7500-5000 yr B.P.) suggest the subseque nt development of moister conditions, which have continued to the pres ent. Temporal patterns in Picea treeline position during the Holocene are different in northern Alaska, northwestern Canada and central Cana da. In Alaska, the treeline is currently at its furthest Holocene exte nt. In northwestern Canada the treeline has retreated since the mid-Ho locene, and in central Canada a mid-Holocene advance and retreat occur red. Such differences between regions suggest a complex response of th is taxon to climatic controls and underscore the need to interpret veg etation changes in relation to regional-scale climate phenomena. If th e pollen chronology is calibrated with reference to U-Th dates, the re lationship of vegetational events to the insolation maximum at ca. 10, 000 yr B.P. is significantly altered. The changed chronology also affe cts the interpretation of pollen records in relation to GCM simulation s, which are partly keyed to calendar years.