| Article en page(s) : |
p. 33 (suppl. 1) : 399-423 |
| Résumé : |
Two experiments were set up to investigate how to maintain or create genetic diversity in artificial or managed populations of plants. Using Arabidopsis thaliana, we established 18 metapopulations of 20 populations each, all with the same initial genetic composition. We tested the effects of the population size, the artificial selection regime and the extinction/recolonisation regime. We report the results of the first four generations of evolution for a trait under selection (precocity) and for allozyme diversity. As expected, overall diversity decreased in each metapopulation, and differentiation among populations increased. As expected, the differentiation was weaker for larger population sizes and in the treatment with extinction and recolonisation with no bottleneck. Artificial selection was effective because the life cycle duration was much reduced. However, most of the reduction occurred during the first generation. We observed an increase of one allele at the LAP-2 locus in all metapopulations, breaching neutral assumptions for this locus. Finally, the selection regime made little difference for small population sizes, whereas large metapopulations were more differentiated when artificial selection was heterogeneous among populations. Altogether, our results agree with theoretical expectations, and provide some new results, which could not have been anticipated. In particular, the overall decrease in genetic diversity was very large (of the order of 20% in 4 generations) even for metapopulations of 2000 individuals. |
Evolution of genetic diversity in metapopulations : Arabidopsis thaliana as an experimental model in Genetics Selection Evolution, 33 (suppl. 1) (2001)
