Although this study was quite interesting, several limitations have to be highlighted. First, the physico-chemical analysis of soils is quite confusing. For example, in the manuscript, they mainly talk about clay fraction and clay content. But their analysis is based on small-size fraction (<62 micrometers) as they explicitly indicated in the Results part. This fraction can also contains silt elements, so, they can't do this kind of shortcut, or at least, they have to clearly indicated it in the M&M part. And they have to homogenize their text (clay vs mud). Second, regarding the sequence analysis part, many details are not given, although it can really impact the results. Indeed, given the large variation and lack of universally accepted standards for microbial profiling studies, additional details should be provided in the bioinformatics sequence analysis like quality metrics applied to remove low quality reads? (E.g. homopolymer, average quality score, chimeras, homogenization step, etc). But, after careful reviewing and the fact that authors took into account all comments, and more particularly on the homogenization step and other analyses, I decided to endorse this publication, considering its impact and interest for biodiversity of arenized soils.