The Arctic has often been regarded for vascular plants and probably also for other groups – as a ”sink”, recruited almost entirely by immigration. Some researchers (see Böcher et al.) even ask for evidence of in situ evolution for all truly Arctic species in the parts heavily glaciated during the Pleistocene. There are exceptions. The largest and probably most important is the now well documented refugium on both sides of the Bering Strait, where large land areas have been continuously ice-free since Tertiary times and where perhaps some hundreds of exclusive species occur. This area has, at least since the 1930s been regarded as some type of cradle or source area for much of the Arctic flora. It has been assumed that plants have spread from Beringia in both directions, through North America and North Asia onwards to a secondary meeting area of the most widespread ones in the North Atlantic in Greenland, Svalbard and northern Fennoscandia. Diversification has taken place during migration so that the plants meeting from east and west in the North Atlantic region are often fairly different even if they have the same origin.

A secondary and obviously much less important source area of Arctic innovation may be the presumed ice-free High-Arctic parts of northernmost Canada (especially parts of Ellesmere Land and Devon Island) and Greenland (parts of Peary Land). Only a very few plants from this area seem to have reached outside Arctic America, probably because they are High-Arctic and not adapted for life in more southerly latitudes. A third possible, but much more disputed refugial source area, are the mountains of West Greenland, Baffin Island, Labrador, Newfoundland and Gaspé Peninsula where there may have been nunataks with plant life.

Emigration from such refugial areas, circumpolar dispersal, and immigration from more southerly source areas such as the western and eastern Cordilleras of North America are assumed to be the origins of the greater part of American Arctic flora. Much emphasis has been placed on species whose distribution is restricted to the possible refugia, i.e. endemic species. Whatever such species may have to say about conditions inside the refugia, they say very little about what may have happened elsewhere, in places where they do not occur.

Recent advances in molecular methods in genetics offer an opportunity for studying the spread of genes rather than entire species and for focusing on more widespread species. It is assumed (and probable) that genetic diversity is highest in or close to the refugia and that a restricted set of genotypes has been able to spread widely through the Arctic. The mapping of genotypes within and outside the refugia may therefore throw light on some of the patterns involved in the migration of plants through the Arctic after the last – Wisconsin or Weichselian – glaciation, and reveal something about how, when (and why) the areas acquired their vascular plant cover.

During Tundra Northwest 1999, live plants, seeds, leaf material samples and dried specimens of several widespread species and species groups were collected throughout the Canadian Arctic. The most important study groups were in the genera Draba, Minuartia, Papaver, Potentilla, and Saxifraga, all widespread and all with a complicated taxonomy.

The use of dried plant material for DNA extraction has increased rapidly in later years. This offers opportunities for the much more time-effective collection of material in the field. For most sites visited in 1999 a more or less complete set of specimens was collected for all vascular plant species found. It is therefore possible to extend the genetic investigations to a much larger range of species, if supporting evidence is required for patterns found in the main study groups.

The data collected also make it possible to evaluate the influence of the different postulated sources in the local floras of 18 well spread areas throughout the Canadian Arctic. This is valuable data for the ongoing Panarctic Flora Project (PAF) which includes in its aims the establishment of a consensus among North American, Russian and West European botanists concerning the phytogeographical division of the entire Arctic, in species concept and nomenclature. Most of the phytogeographical dispute has been about how to di vide the Canadian Arctic. Russian proposals recognize only small western and south-eastern areas as Beringian and Atlantic, in a floristic meaning, whereas the majority are less clearly divided. The results from 1999, in addition to much previous information, support recognition of an extended Beringian area in the west, reaching east to include parts of Banks and Victoria Islands and the mainland coast east to Cape Bathurst, and also a much extended Atlantic area in the east including the surroundings of Hudson Bay and most of Baffin Island.

The Beringian ”revision” in the west is the result of the identification of an element of species typical of parts of the western islands and the mainland coast, but absent in the central and eastern parts of the Canadian Arctic and also generally in the mainland mountains in Yukon Territory and northern Alaska. This element might be an American counterpart to the impressive element in eastern and northern Chukotka on the Russian side, especially on Wrangel Island. On the Russian side, this element is interpreted as remnants of the Pleistocene Beringian flora of the shelf areas which were dry land and connected Northeast Asia and North America during much of the glaciations. A similar interpretation is natural for the American counterpart.