Population genetics of cultivated common buckwheat, Fagopyrum esculentum Moench. IX. Concluding remarks on worldwide survey of allozyme variability

By adding data on allozyme variability in 35 populations from southern and central China, I completed a worldwide survey of allozyme variability in common buckwheat. The following remarks were born out from the survey. (1) At all loci, allele frequency does not vary so much among the populations from a wide range of Asian countries. (2) Most of the evolutionary events which might occur during the spread of buckwheat cultivation are the losses of variant alleles by random drift. (3) Drastic changes of gene frequency have been observed only in marginal populations, in Kumaun, Garwhal hills and Kashmir in India and in southern Europe. (4) Cultivated buckwheat has more genetic variability than the natural populations of the wild ancestor. Large population size and complete panmixis of a population and enough migration between populations may be responsible for observation (1) and may also have led to the accumulation of variant alleles in cultivated populations, i.e. observation (4). Random drift, such as founder effect, took place only in marginal populations, accounting for observations (2) and (3). Buckwheat provides an example that contradicts Vavilov's assertion; it says that the center of genetic diversity of a cultivated plant is the place of its origin. Buckwheat is apparently monocentric, but the pattern of variation suggests domestication over a wide area, instead of progressive decline in diversity from the center of origin to the periphery. The structure of buckwheat populations and their mating system are mainly responsible for this unexpected result.