Genetic diversity at 11 loci encoding nine enzymes was studied in 23 populations of Japanese beech Fagus crenata Blume distributed throughout the range of the species. Levels of genetic diversity were high for both within species (expected mean heterozygosity: 0.194) and within populations (expected mean heterozygosity: 0.187), whereas the level of genetic diversity among populations was low (GsT = 0.038), as observed in various long-lived, woody plants. Despite the low differentiation among populations, geographical patterning of the variation was observed. Populations in southwestern Japan tended to have greater within-population variation and to be more highly differentiated when compared with those in northeastern Japan. In addition, allele frequencies observed at eight loci were significantly related to latitudinal and/or longitudinal gradients and showed clinal variation across the range of the species. Principal components analysis revealed that the populations tended to cluster according to their geographical locations. The nonrandom patterns of variation were probably shaped by relatively recent historical events such as late-Quaternary migration and founding events.

Genetic diversity at 11 loci encoding nine enzymes was studied in 23 populations of Japanese beech Fagus crenata Blume distributed throughout the range of the species. Levels of genetic diversity were high for both within species (expected mean heterozygosity: 0.194) and within populations (expected mean heterozygosity: 0.187), whereas the level of genetic diversity among populations was low ( GsT = 0.038), as observed in various long-lived, woody plants.Despite the low differentiation among populations, geographical patterning of the variation was observed. Populations in south-western Japan tended to have greater within-population variation and to be more highly differentiated when compared with those in north-eastern Japan. In addition, allele frequencies observed at eight loci were significantly related to latitudinal and/or longitudinal gradients and showed clinal variation across the range of the species. Principal components analysis revealed that the populations tended to cluster according to their geographical locations. The nonrandom patterns of variation were probably shaped by relatively recent historical events such as late-Quaternary migration and founding events.