Soil microbiome associations with tillage differ by organismal group, community metric, and sensitivity to sampling imbalance

Agricultural management practices alter soil microbiomes, but the magnitude and nature of the associations can differ by which microbial groups and community metrics are examined. Many studies have examined either prokaryotes or fungi in soils from agricultural experiment plots. Here, we assessed prokaryotes (bacteria plus archaea), fungi, and arbuscular mycorrhizal fungi (AMF) from on-farm soils under soybean (Glycine max (L.) Merr.) in Ohio at two depths (0–10 cm and 10–30 cm). Microbiomes were characterized via amplicon sequencing targeting regions of the 16S rRNA gene for prokaryotes, internal transcribed spacer (ITS) region for fungi, and 18S rRNA gene for AMF, and we evaluated tillage associations for community diversity, composition, management-associated amplicon sequence variants (ASVs), and lineage correlation networks. Community alpha diversity metrics of richness and Shannon diversity did not differ significantly with tillage management. Management explained 4.7%, 6.5%, and 3.1% of Bray–Curtis variation in 16S, ITS, and AMF communities, respectively; none of the profile-restricted PERMANOVA tests met the adjusted significance threshold (q = 0.0555, 0.3770, and 0.3018). Differential-abundance models identified 20 management-associated ASVs at Benjamini–Hochberg (BH) q ≤ 0.05, comprising 5 prokaryotic, 14 fungal, and 1 AMF ASV. These ASVs included an ASV assigned to Clonostachys rosea, a species containing strains with documented biological control activity. However, this activity was not evaluated in this study. Correlation-network topology differed between groups, but several contrasts changed substantially after balancing the sample numbers. These results describe management-associated patterns, and their interpretation depends on the analytical summary, sample balance, and confounding with collection date and site.