Presently, there is considerable debate whether variation in microbial communities over geographic distance results from environmental filtering, or if historical contingencies and geographic barriers also contribute to spatial structure in community composition. We address this debate by examining the spatial patterns of a deeply diverse and divergent phylum, the Actinobacteria, in a northern hardwood forest ecosystem. Our four field sites occur along a chronosequence of ecologically and edaphically matched sugar maple stands which established 9,500 to 14,000 years following glacial retreat in Michigan, USA. If dispersal limitation is not a factor structuring these communities, then similar communities of Actinobacteria should occur in all sites, and variation in community composition should be structured by environmental factors. However, if long-distance dispersal is uncommon for Actinobacteria, as glaciers retreated from south to north, new actinobacterial species would accumulate at each site at different rates; therefore, different communities should occur along our chronosequence. To test this hypothesis, we characterized actinobacterial communities using T-RFLP fingerprints, along with cloning with sequencing, of actinobacteiral 16S rRNA genes from the surface soil of ecologically similar forest stands that differ in time since glacial retreat. Taxonomic, phylogenetic, and multivariate statistical analysis were used to evaluate community similarity, the distance-decay relationship, and the relatedness of community patterns to environmental variation, climactic factors, and geographic distance.
Results/Conclusions
At every level of examination, actinobacterial community composition more closely correlated to geographic distance than either environmental or climactic characteristics. We observed a significant distance-decay relationship based on T-RFLP results (P = 0.02, slope = -0.09), providing evidence for dispersal limitation in actinobacterial communities. However, phylogenetic analysis revealed that dispersal limitation may not be the only ecological factor shaping these communities, due to significant amounts of unique lineages at the oldest and youngest sites, and the lack of communities clustering along the phylogenetic tree. Nevertheless, on many levels of resolution, we have evidence that geographic distance, a surrogate for time, leads to decreased community similarity in the presence of environmental homogeneity across this chronosequence of forested sites.