Reconstruction of a beech population bottleneck using archival demographic information and Bayesian analysis of genetic data
Résumé
A major expected consequence of climate change is altitudinal and latitudinal shifts in the geographic ranges of plant species. These shifts and the associated demographic fluctuations have been shown to have pervasive consequences for genetic diversity. For instance, imprints of postglacial recolonisation events on genetic structure can still be observed after millennia and at the continental scale. However, explicit empirical tests of how specific contraction/expansion processes may shape patterns of genetic diversity remain scarce. In particular, there are few studies addressing these processes at the scale at which dispersal and expansion actually take place and at which management guidelines are required, i.e. at the local scale and over several generations. This study takes advantage of a unique opportunity to investigate recolonisation dynamics and their consequences for genetic diversity in European beech ([i]Fagus sylvatica[/i] L. Crantz) on Mont Ventoux, France, at an intermediate spatial and temporal scale employing a combination of historical records and genetic data. From the 12th century to the mid-18th century human activities reduced the beech forest on Mont Ventoux to a few remnant stands, whereupon forest management policy and practices changed to favour expansion of beech forest. Analyses of historical forest maps and management documents showed that beech refugia were not as scarce as has been reported. Since 1876 beech forests have expanded by as much as 65%, moving at an average rate of 27 to 38 m/year. Current patterns of genetic diversity and differentiation were investigated using 1825 individuals in 51 plots from three regions on Mont Ventoux genotyped at 13 microsatellite markers. The strongest differentiation was between plots (Fplots/total = 3.5%), with a surprisingly low level of differentiation between zones (Fzone/total = 0.9%). Isolation by distance was detected up to 750 m. Approximate Bayesian computation (ABC) method for estimating the probabilities of demographic scenarios based on microsatellite data supported the hypotheses of population expansion during the last 200 years, and divergence between the populations on the north and south faces of the massif, consistent with historical data.