OTU generation methods captured consistent community patterns across soil conditions
All three OTU generation methods consistently demonstrate that total soil microeukaryotic community composition clearly differentiated based on soil condition (wet or mesic-dry), but no pattern was detected in relation to the presence of F. meleagris , as observable in nMDS ordinations (Fig. 3). The observed separations were statistically significant (p =0.001) as indicated by a marginal PERMANOVA test (Table S4) and remained when the importance of rare OTUs was down-weighted by square root transformation of relative abundances (Fig. S6, Table S5). Similar to observations for the plant community (Fig. 2), microeukaryotic community composition was more variable among samples in wet soil conditions compared to mesic-dry soil conditions (Fig. 3). Mesic-dry samples clustered closer together indicating that communities were more similar across samples (Fig. 3). When analyzing fungal and protist community composition separately, we observed the same significant separation (p <0.001) based on soil conditions (wet or mesic-dry), but not in relation to the presence ofF. meleagris (Fig. S7, Table S6). While still significant, the separation is visually less distinct for the protist community based on OTU_S and OTU_C (Fig. S7 d, f) compared to OTU_A (Fig. S7b). The tight clustering of samples from mesic-dry conditions is recovered in both fungal and protist communities (Fig. S7).
Overall, phylum-level taxonomic composition was also comparable across the three OTU generation methods used (Fig. 4). Based on read abundance, protist communities were dominated by the Ciliophora (Alveolata) in both wet and mesic-dry soil conditions (Fig. 4a). The relative abundance of Alveolata was slightly lower when communities were characterised using OTU_C and OTU_S compared to the OTU_A dataset (Fig. S8). The Rhizarian phyla Endomyxa, Phytomyxea, and Filosa were also observed in both conditions and were more abundant when reads were clustered into OTU_C and OTU_S compared to OTU_As (Fig. 4a). In wet conditions, a larger proportion of reads within both Alveolata and Rhizaria could not be identified at the phylum level highlighting the potential for future studies of poorly known lineages at this site. The proportion of Ciliophora was smaller in wet vs mesic-dry conditions. In both soil conditions, Ascomycota was the most common fungal phylum, and together with Basidiomycota, made up over half of the sequenced fungal community in wet soil conditions (Fig. 4b). In mesic-dry soil conditions on the other hand, Mortierellomycota made up a larger fraction of the reads, around 30%. Sequences assigned to Glomeromycota, which encompass all arbuscular mycorrhizal fungi, were rare at this site, despite known high abundance based on spore counts (personal observations of S.E.S.). Chytridiomycota were also more abundant in mesic-dry compared to wet soil conditions, while Rozellomycota made up around 10% of the reads in both conditions. Close to 20% of fungal OTUs remained unidentified at phylum level across all three methods (Fig. 4b). Many of these unidentified lineags cluster with Zoopagomycota, Kickxellomycota and Rozellomycota in the fungal tree (Supplementary datafile 2).