Search

It has been suggested that many arctic-alpine plant species have limited dispersal ability and cannot have arrived in Scandinavia and the arctic archipelago of Svalbard by long-distance dispersal after a total glaciation. It has therefore been proposed that such species must have survived the entire glaciation(s) in ice-free refugia in southern Norway, northern Norway and Svalbard. We investigated random amplified polymorphic DNA (RAPD) variation among 28 populations from Norway and Svalbard of one of these arctic-alpine ‘short-distance dispersers’, the selfing polyploid Saxifraga cespitosa. In an analysis of molecular variance (AMOVA), more variation was found among populations within the three postulated refugia regions (45%) than among these regions (25%). Spatial autocorrelation (Mantel) analyses showed that the genetic distance monotonously increased with increasing geographical distance. In UPGMA and PCO analyses, the populations from Norway and Svalbard formed a south–north cline that continued across the Barents Sea barrier. The results suggest that there has been recent dispersal among the three postulated refugia regions and thus that postglacial dispersal into these refugia regions from other distant areas also must represent a possibility. The observed geographical pattern of the genetic variation may have been established after expansion from different source areas outside the North European ice sheet and/or from different refugia areas ‘within’ the ice sheet, but it is probably not possible to distinguish among these alternatives. The results for S. cespitosa are consistent with a dynamic late- and postglacial scenario with extensive plant dispersal, and support the conclusion from our previous study of the outbreeding Saxifraga oppositifolia; the hypothesis of glacial survival in Norway and/or Svalbard is superfluous. [ABSTRACT FROM AUTHOR]

The arctic-alpine Saxifraga oppositifolia has recently been suggested to have survived the last glaciation in high-arctic refugia, based on a finding of more genetic (RFLP) varaition in Svalbard compared with more southern areas. To elucidate the migration history of this allogamous species, we analysed 18 populations from Norway, Svalbard and Novaya Zemlya using random amplified polymorphic DNAs (RAPDs). There was no more RAPD variation in the high Arctic than further south. In an analysis of molecular variance (AMOVA), most of the RAPD variation was found within populations (64%). There was less intrapopulational varaition in Svalbard (65%) than in northern Norway (78%) and southern Norway (86%), suggesting that there is more inbreeding towards the north, probably because of lower pollinator activity. Twenty-eight per cent of the RAPD variation was found among populations within these geographical regions, and only 9% was found among the regions. In PCO and UPGMA analyses, plants and populations of different geographical origins were to a large extent intermingled. There was, however, a distinct, south-north clinical geographical structuring of the RAPD variation both in the PCO analysis and in a spatial autocorrelation (Mantel) analysis. These results suggest that there has been extensive gene flow among more or less continuosly distributed populations of S. oppositifolia during the Weichselian, and the extant Nordic population were established after massive, centripetal immigration from these genetically variable, periglacial populations. The postglacial period may not have been sufficiently long for the subsequently isolated populations of this long-lived, allogamous perennial to diverge. Given the high levels of migration inferred from this study, genetic differentiation of glacial survivor populations, if any existed, would most likely have been swamped in the postglacial period. Thus, our molecular data support recent conclusions based on palaeobotanical and biogeographical data that the glacial survival hypothesis is superfluous. [ABSTRACT FROM AUTHOR]