One-step PCR amplicon sequencing libraries perform better than two-step when assessing soil microbial diversity and community profiles

Despite adoption of high-throughput sequencing of PCR-amplified microbial taxonomic markers for ecological analyses, distinct approaches for preparing amplicon libraries exist. One approach utilises long fusion primers and a single PCR (one-step) while another utilises shorter primers in a first reaction, before transferring diluted amplicons to a second reaction for barcode index incorporation (two-step). We investigated whether transferring diluted amplicons risked creating artificially simplified, poorly diverse communities. In soils from three sites with paired cropland and forest, one-step yielded higher alpha-diversity indices, including detection of two–four times more unique taxa. Modelling expected taxa per sequence observation predicted that one-step reaches full coverage by 104 sequences per sample while two-step needs 105–109. Comparisons of rank abundance demonstrated that two-step covered only 38%–69% of distributions. Beta-diversity showed better separation of communities in response to land use change under one-step, although both approaches showed a significant effect. Driving differences was underestimation of relatively minor taxa with the two-step procedure. These taxa were low in abundance, yet play important roles in carbon cycling, secondary metabolite production, anaerobic metabolism, and bacterial predation. We conclude that one-step amplicon libraries are advisable for studies focussed on diversity or relatively minor yet functionally important taxa.