Chloroplast DNA variation within the Nordic countries

Abstract

Chloroplast DNA (cpDNA) variation was studied in the white oak species, Quercus robur and Quercus petraea, from Denmark, Norway, Sweden and Finland. A total of 474 trees from 91 populations were screened within the region for a set of previously characterised cpDNA haplotypes. Within the sample, haplotype 1, known to have originated from an Italian Pleistocene forest refugium, was the most common (55%), and was well-distributed across all Scandinavian countries and dominated in Sweden. We propose that material of Italian origin was part of the first wave of post-glacial migration of oak into Scandinavia, and that the Italian haplotype migrated into Finland from Sweden across the Baltic Sea. In Finland, a sharply defined transition zone between the haplotype of Italian origin and haplotype 5, originating from an eastern European refugium in the Balkans, was observed in southern Finland, and confirms the findings of a previous study [Heredity 80 (1998) 584]. Evidence from the study of Csaikl et al. [For. Ecol. Manage., this issue] confirms that this haplotype migrated into Finland via an eastern route through the Baltic States. Another haplotype of Balkan origin (haplotype 7) was only common in southern Denmark and known to be present throughout the native German forests [For. Ecol. Manage., this issue], from where we believe it migrated into Denmark. It appears probable that this haplotype reached Denmark later than the Italian lineage as its progress towards north appears to have been hindered by a bottlenecking in central Denmark. Finally, three haplotypes originally of Iberian origin (haplotypes 10-12) were also found at intermediate frequency throughout the Nordic countries sampled, except Finland. This pattern of distribution suggests that migration occurred concurrently or just behind that of the Italian lineage. One of the Iberian haplotypes (haplotype 12) is rare in autochthonous stands, but found at a frequency of 24% in planted non-autochthonous stands in Denmark. Overall, the proportion of populations fixed for a single haplotype was high in natural, autochthonous stands (77%) and lower in artificial stands (54%). In addition, population subdivision was higher (GsT = 0.87) and diversity lower (hs = 0.075 and hT = 0.6) in autochthonous stands than in artificial stands (GST = 0.69, hs = 0.22, hT = 0.72). The pattern of haplotype distribution across Scandinavia and Finland appears fairly clumped and is discussed in relation to natural gene flow, leptokurtic post-glacial dispersal and human influences. Finally, no relation between cpDNA haplotype and six quantitative traits measured on autochthonous stands in Denmark was found. The potential for cpDNA haplotype screening for setting gene conservation and seed certification priorities is discussed. © 2002 Elsevier Science B.V. All rights reserved.