Above and belowground community analysis
Statistical analysis were performed in R using RStudio (RStudio Team, 2015). The community composition of plant species recorded in all plots along transects 1 and 4 was visualized by non-metric Multidimensional Scaling (nMDS) using Bray-Curtis dissimilarities. Both plant and belowground analyses were made with the vegan software package (version 2.5-7; Oksanen et al., 2019). The two-sample I-test was used to test for significant difference in mean gravimetric soil moisture between the two soil moisture regimes (wet vs. mesic-dry). The three OTU occurrence tables were transformed to relative abundances for each sample and used for community analysis. nMDS ordination plots were generated using the ‘metaMDS´function in vegan. To down weight the importance of common taxa, the analysis was repeated using square root transformation of relative abundance data prior to calculation of the Bray-Curtis dissimilarity. Marginal and individual PERMANOVAs were conducted on all three datasets and two standardizations described above, using the ‘adonis’ function in the vegan to test for the marginal and overall effect of soil moisture regime (wet vs. mesic-dry) and presence/absence of F. meleagris on shaping belowground microeukaryotic communities at the study site.
After taxonomic assignment as described above, the three OTU occurrence tables were divided into two separate datasets for fungal and protist communities separately (ie. non-fungal microeukaryotes). The ordination and PERMANOVA tests described above were repeated for these taxonomically distinct communities.
Distribution and abundance of OTU_Ss were visualized across the contrasting soil conditions wet vs. mesic-dry soil conditions and presence/absence of F. meleagris using a Venn-diagram (Heberle et al., 2015). The relative abundance of unique and shared fungal and protist OTU_Ss were calculated across samples for the two contrasting conditions. To identify differentially abundant taxa in contrasting soil moisture regimes (wet vs. mesic-dry) and presence/absence of F. meleagris , the ‘phyloseq-to-deseq´ function in the phyloseq package (v 1.34.0; (McMurdie & Holmes, 2013) was applied to OTU_S occurrence tables separately. The generated phyloseq object was analysed using the DESeq2 tool (DESeq package version 1.30.1; Love et al., 2014). Identified taxa and their differential abundance were illustrated using the ggplot2 R package (Wickham, 2016).