Genetic legacies of historic landscape modifications on American Beech populations

Background/Question/Methods

Throughout eastern US, extensive selective logging and landscape-level forest clearance for agriculture have occurred since the 1600s, shaping forest structure. Though these disturbances are fairly recent relative to the lifespan of American beech (Fagus grandifolia), they seem likely to have genetic effects. The importance of these disturbances on forest genetics will be, in part, a function of other features such as population isolation, location (central or peripheral) and historical abundance. Here, we examine the levels, patterns and magnitude of influence of land-use relative to other processes on genetic diversity and structure of beech, a wind-pollinated and disturbance-sensitive species. We asked what are 1) the magnitude of genetic diversity loss due to land-use, and 2) the relative influence of population isolation, location and abundance on this genetic diversity. We collected 500 individuals from 28 sites across Massachusetts and genotyped them at eight highly variable microsatellite loci. We grouped populations by 1) forest type – primary (selectively logged forests) vs secondary (forests that naturally regenerated from abandoned agriculture lands) and by 2) location - inland (central populations) vs coastal/island (peripheral and more isolated populations). Genetic diversity, genetic differentiation and levels of gene flow were calculated and compared between groups and among all populations.

Results/Conclusions

Population genetic diversity estimated from multilocus genotypes indicated high levels of observed heterozygosity among all populations (ranging from 0.55 to 0.80). The overall pattern of genetic diversity (measured by allelic richness A_R, observed heterozygosity H_o, and within population gene diversity, H_s) indicated a significant loss in secondary (A_R=6.042, H_o=0.502, P<0.05) compared to primary forests (A_R=6.819, H_o=0.580, P<0.05).  This pattern was more evident when comparing secondary and primary forests in coastal/island group than the inland group. We examined the extent of genetic connectivity among populations in both groups and found strong genetic differentiation among coastal populations (F_st=0.162, P<0.0001), spanning a distance of ~108 km.  High levels of gene flow were inferred from inland populations (F_st=0.046, P<0.001), suggesting a single panmictic population even with a distance of ~106 km. These results suggest that forest connectivity (i.e. high gene flow may counteract loss of genetic diversity) and population location is important in determining the extent of genetic diversity reduction due to land-use. Isolated and peripheral populations where population sizes are small (i.e. coastal populations) may be more vulnerable to human land-use than those near core of range i.e. inland populations. This has important implications to future forest conservation and management strategies.