Beringia – a refugium for arctic and boreal biota?

The present-day arctic flora comprises approximately 1 500 species and is of relatively recent origin (Murray, 1995). Throughout most of the Tertiary (65–2 Ma), forests grew at high latitudes in the Arctic (Murray 1995, McIver and Basinger, 1999) and tundra did not appear until the late Pliocene (Matthews and Ovenden, 1990). Initially tundra was distributed discontinuously, but a Circum-Arctic belt was present by 3 Ma (Matthews, 1979). Little is known of the origins of arctic plants, although it is supposed that many such plants are derived from ancestral stocks which occurred on high mountains from the south of Asia, as well as Europe and North America (Hultén, 1937, Tolmachev, 1960, Weber, 1965, Hedberg, 1992, Murray, 1995). These mountains form part of ranges connected to the Arctic, along which plants could have migrated northwards as global temperatures dropped significantly from the mid-Miocene onwards (Lear et al. 2000, Zachos et al., 2001). In addition some arctic plants may be descended from shrubby and herbaceous elements of the Tertiary arctic forests that occupied open bog, riparian and well-drained upland habitats in the Arctic during the late Tertiary (Murray, 1995). In the Quaternary (approximately 2 Ma until present), the distribution and composition of the arctic flora was greatly affected by the advance and retreat of ice sheets. Traditionally it was believed that during glacial periods all northern areas were covered by ice to a similar extent and that arctic animals and plants migrated southwards of advancing ice-sheets to survive in southern refugia (Darwin, 1859, Hooker, 1862). However this belief was challenged in 1937 by the Swedish botanist Eric Hultén, in his book Outline of the History of Arctic and Boreal Biota during the Quarternary Period. Hultén drew on geological evidence and a vast body of his own phytogeographical evidence to propose that most of Northeast Russia and Northwest America (Alaska and the Yukon) remained ice-free during Quaternary glaciations and served as a massive northern refugium for arctic and boreal biota. Hultén called this region Beringia. Today it is defined as the area between the River Lena (125°E long.) in northeastern Siberia to the east, the River Mackenzie (130°W long.) in northwestern America to the west, the Arctic Ocean to the north and southern Alaska and the middle Kuril Islands to the south (Abbott and Brochmann, 2003).

Beringia served as a land-bridge between Eurasia and North America throughout the Tertiary until approximately 5 Ma when it was severed by the formation of the Bering Strait (Marincovich and Gladenkov, 1999, 2001). During the Quaternary, the land bridge reformed during major glaciations when sea levels fell by 100–135 m (Hopkins 1973, Clark and Mix, 2002). Hultén (1937) further proposed that many arctic plants obtained a circumarctic distribution early in the Quaternary period. However during each subsequent glaciation large parts of their distributions were destroyed, only to reform during interglacials through recolonization of deglaciated areas from Beringia and from southern regions. Some species would have been less successful than others at migrating back into these areas and would retain a fragmented distribution, with large gaps occurring between geographically disjunct areas.

Botanists on the Beringia 2005 expedition have sampled the Beringian flora – from Kamchatka in the south to Wrangel Island in the north on the Russian side, and from Nome to Barrow on the American side – in order to test Hultén’s proposals on the history and evolution of the arctic flora in the light of new phytogeographical and molecular evidence. Among the plants specifically selected for study are Alnus, Cassiope, Dryas (picture 1), Empetrum, Papaver section Meconella, Potentilla (picture 2), Saxifraga section Mesogyne, Vaccinium uliginosum s. lat., selected species groups of the large genera Cerastium and Draba and many additional groups. Among the plants are trees as well as dwarf shrubs and herbs. One of the interesting aims of this collaborative research is to contrast the historical biogeography of different life forms.

The fieldwork

Collecting plants is a relatively uncomplicated matter in the field, but preservation of the material for the future, either as herbarium sheets or as a leaf sample for molecular analyses, requires careful processing. Elven and Solstad went to the northern sites in Chukotka and Alaska on leg 2A while Eriksen and Andersson visited Kamchatka on leg 2C. The icebreaker Oden formed a mobile platform for parts of leg 2A whereas leg 2C was based entirely on campsites. Hence the technical facilities for plant processing differed somewhat between the teams. During daily hikes on the selected sites, plants were collected and kept separate and fresh in plastic bags. At the end of the day plants were catalogued and put into plant presses for drying. Before Oden arrived, air drying was the only possibility in Chukotka, whereas on Oden the presses were placed in a drying cupboard. The Kamchatka team, lacking electricity at the base camps, used purposely designed mobile drying facilities (picture 3). The boxes were heated by multi-fuel kitchens and could be compacted when transported by truck or helicopter. Material for molecular analyses was picked from each voucher and quick-dried in silica gel.

Preliminary results

In Kamchatka four sites were collected: Ust’-Bolsheretsk on the southwest coast of the peninsula, Ichinskaya volcano in Central Kamchatka, Karaginsky Island off the coast in the northeast, and finally Mutnovskaya volcano south of Petropavlovsk-Kamtchatskiy in the southeast. At all sites we were able to sample plant material from some of the target species. In total 429 specimens were collected, of which 92 were preserved on silica gel for molecular analysis.

A recent paper summarizing pollen records from the Palaeoenvironmental Arctic Sciences Database into spatial-temporal patterns concludes that Alnus, as well as other species of trees (Picea, Pinus, Larix, and Betula), survived within the Beringian refuge during the Last Glacial Maximum (Brubaker et al., 2005). This result challenges the traditional view of Beringia as a treeless steppe. In Göteborg cpDNA microsatellite data on material from Alnus will be used during the spring of 2006 to establish which of the above-mentioned hypotheses is corroborated by patterns of haplotypes. Parallel studies by quaternary geologists also participating on the Beringia 2005 expedition in Kamchatka – Bennett et al. – on stratigraphic data from lake sediment cores will hopefully provide additional information on past vegetation changes to compare our data with.

In Chukotka five full sites were covered in Penkigney Bay (2 sites), Lavrentia Bay, the Kolyuchin Bay surroundings and Wrangel Island. In Alaska two site regions were covered around Barrow and in Seward Peninsula. Due to logistic problems only two sites were supported by Oden. In total 1 378 specimens were collected, of which 470 were preserved on silica gel for molecular analysis. We made a special effort to collect samples for molecular analysis in Wrangel Island as the hypothesis of the importance of this island in Beringian biogeography has been unsupported until now by molecular evidence. The material collected (both herbarium specimens and leaf samples) is already in use for numerous studies in phylogeography and systematics.