Fig. 2

16S primer choice affects sequencing quality and diversity measures but not community-level results. A Read quality. Earth and MiSeq primers yielded reads with high sequencing quality, whereas primer set 27F/534R yielded lower read quality and was excluded from downstream taxonomic analyses. Between sample types, read quality was comparable for each primer set. B Richness. Earth primers yielded higher alpha richness based on ASVs across all sample types. In farm A, biofilter water featured the most ASVs, followed by tank water and, finally, tank biofilm. Richness in farm B's biofilm was as high (Earth) or higher (MiSeq) than farm A's biofilter water. C Diversity. Samples amplified with Earth primers displayed higher alpha diversity (Shannon index) than samples amplified with MiSeq primers. Patterns are overall similar to richness (panel B). The similar diversity patterns suggest that it would be possible to compare community studies using different primers at the relative scale. D Community composition. The biological and technical replicates (samples 2, 20–21) were highly similar in composition, suggesting that primer selection does not impact spatio-temporal findings at the phyla level and indicates that reproducible results can be obtained with short amplicon sequencing. Indications for succession can be seen in the tank biofilm samples, with increasing complexity from young to older biofilm. Finally, Farm B's tank biofilm samples resembled Farm A's biofilter water samples, suggesting that microbial communities with RAS become similar in complexity over time, potentially reaching a stable, mature state. Abbreviations after the phylum name indicate that the phylum was detected in other platform datasets; PB = PacBio, MG = Metagenomics