Insight into the microbial community of denitrification process using different solid carbon sources: Not only bacteria.

• Novel carbon source and corncob had better denitrification efficiency than polycaprolactone and polyvinyl alcohol-sodium alginate. • The taxonomical composition of bacterial, fungal and archaeal communities can be altered by types of solid carbon source. • Some fungi and archaea species were conducive to solid-phase denitrification (SPD) systems. • Co-occurring network showed some potential collaboration between bacterial fungal and archaeal communities in SPD systems. • Amplicon sequence variant (ASV) can identify more taxonomies that operational taxonomic unit (OTU) cannot in this study. There is a lack of understanding about the bacterial, fungal and archaeal communities' composition of solid-phase denitrification (SPD) systems. We investigated four SPD systems with different carbon sources by analyzing microbial gene sequences based on operational taxonomic unit (OTU) and amplicon sequence variant (ASV). The results showed that the corncob-polyvinyl alcohol sodium alginate-polycaprolactone (CPSP, 0.86±0.04 mg NO 3 −-N/(g·day)) and corncob (0.85±0.06 mg NO 3 −-N/(g·day)) had better denitrification efficiency than polycaprolactone (PCL, 0.29±0.11 mg NO 3 −-N/(g·day)) and polyvinyl alcohol-sodium alginate (PVA-SA, 0.24±0.07 mg NO 3 −-N/(g·day)). The bacterial, fungal and archaeal microbial composition was significantly different among carbon source types such as Proteobacteria in PCL (OTU: 83.72%, ASV: 82.49%) and Rozellomycota in PVA-SA (OTU: 71.99%, ASV: 81.30%). ASV methods can read more microbial units than that of OTU and exhibit higher alpha diversity and classify some species that had not been identified by OTU such as Nanoarchaeota phylum, unclassified_ f_ Xanthobacteraceae genus, etc., indicating ASV may be more conducive to understand SPD microbial communities. The co-occurring network showed some correlation between the bacteria fungi and archaea species, indicating different species may collaborate in SPD systems. Similar KEGG function prediction results were obtained in two bioinformatic methods generally and some fungi and archaea functions should not be ignored in SPD systems. These results may be beneficial for understanding microbial communities in SPD systems. [Display omitted] [ABSTRACT FROM AUTHOR]