To better understand the factors that influence the distribution of bacteria associated with mosses, the communities inhabiting in five moss species from two different habitats in Beijing Songshan National Nature Reserve were investigated using the high-throughput sequencing method. The sequencing was performed based on the bacterial 16S rRNA and 16S rDNA libraries. Results showed that there are abundant bacteria inhabiting in all the mosses sampled. The taxonomic analysis of these bacteria showed that they mainly consisted of those in the phyla Proteobacteria and Actinobacteria, and seldom were from phylum Armatimonadetes, Bacteroidetes and Firmicutes. The hierarchical cluster tree, based on the OTU level, divided the bacteria associated with all samples into two branches according to the habitat types of the host (terrestrial and aquatic). The PCoA diagram further divided the bacterial compositions into four groups according to both types of habitats and the data sources (DNA and RNA). There were larger differences in the bacterial community composition in the mosses collected from aquatic habitat than those of terrestrial one, whether at the DNA or RNA level. Thus, this survey supposed that the habitat where the host was growing was a relevant factor influencing bacterial community composition. In addition, the bacterial community detected at the RNA level was more sensitive to the habitat of the growing host, which could also be proved by the significantly differences in the predicted function by PICRUSt and the metabolically active dominant genera between different groups. This study expands the knowledge about the interactions between mosses and microbes.
Keywords: Aquatic and land mosses; Bacterial composition; High-throughput DNA sequencing.