Genetic variation within and among naturally regenerating populations of alder (Alnus glutinosa)

To assess the inter- and intrapopulation genetic variation in the filial generation (F1) of alder (Alnus glutinosa (L.) Gaertn.), 11 naturally regenerated populations were analysed. Their parental populations (P), represent the whole Polish territory and belong to three phytosociological associations with alder: typical alder swamp forest Carici elongatae-Alnetum (Ce-A); alder riparian forest Circaeo-Alnetum (C-A); and ash-elm riparian forest Fraxino-Ulmetum (F-U). F1 populations are grown in a common-garden experiment (provenance trial). Genotyping of individual trees has been carried out by analysis in a bud tissue allele frequency in the 21 isozyme putative loci of 10 enzymes. Differences between populations in respect to the level of genetic diversity were not high. Genetic diversity measured as the number of effective alleles per locus was the highest (Ne = 1.65) in population Wińsko originating from F-U (where also the inbreeding coefficient was the highest, F = 0.429), and the lowest (Ne = 1.48) in population Sławki from Ce-A. In all investigated populations, observed heterozygosity (Ho = 20%) was lower than expected from H-W equilibrium (He = 29%). The highest genetic variation expressed as percentage of polymorphic loci (77.3%) was observed in the offspring populations from Ce-A, and the smallest (69.9%) in the populations originating from F-U. It seems that the low genetic differentiation between populations is probably connected with long-distance seed dispersal via river systems. Alder seed can be transported over long distances thanks to periodical flooding. There is some gene flow between alder populations, with about 2.5 immigrants successfully entering a population per generation (Nm = 2.55). The level of population subdivision within A. glutinosa was low (Fst = 0.089). There was no significant genetic differentiation between populations from different phytosociological associations. Mantel test exhibited no significant correlation (r = 0.077) between genetic and geographic distance. In the dendrogram constructed according to Nei (1972) on the basis of interpopulation genetic distances, many small groups can be observed