Demarcation of species can never be fully done for Procaryotes as they do not constitute Menderian population. Thus we need to employ Operational Taxonomic Unit (OTU) defined by the threshold of 97 or 98% similarity in 16S rDNA for identification of Bacteria and Archaea. Though it was suggested that roughly 6000 "species" existed in a given forest soil of cm³ which contains 4.8 x 10⁹ cells (Torsvik et al. 2002) or far more than that (Gans et al. 2005), I herein try to depict the difference of soil layers of A and B horizons from the difference in microbial constituents revealed by 16S rDNA.

It is astonishing that B-horizon contained more diversified Bacterial constituents than that of A-horizon. And a fingerprint technique showed that Bacterial constituents of B-horizon became similar to that of in situ A-horizon by nutrient enrichment cultivation of several weeks. In addition, a fact that a soil animal Parafontaria laminata may control dominant microbes through its feeding behavior was suggested from another experiment.

When look at the function of microbes in nitrogen cycling in a forest system a hot spot of denitrification was found at the bottom of a watershed, where the water content was as high as 70%. Functional gene analysis of the candidate bacteria for denitrification there showed "high activity was supported by highly diversified microbial community".