*2.7. Statistical Analysis*

The Statistica 13.1 package [66] was used for statistical analyses. Homogenous groups were determined with the Tukey's test, at p = 0.05, and respective results were presented graphically using principal component analysis (PCA) and graphs categorized for dependent variables (category X) and the grouping variable (category Y). Using the analysis of variance (ANOVA), F and P values were calculated for all parameters tested (Table S1). Relative abundance of microorganisms in soil samples was visualized with the use of STAMP 2.1.3 software, using a two-way test for statistical hypotheses: G-test (w/Yates') + Fisher's and Asymptotic with CC confidence interval method [67]. Genomic data were presented in the circular system using Circos 0.68 package [68]. Visualization of relative abundance data was performed using sequences with contribution exceeding 1%. The read-outs below 1% were summed up with the other nonclassified ones in a sample. To determine bacterial diversity at the level of each taxonomic group, Shannon–Wiener (H) and Simpson (D) indices were calculated using all metagenomic data. In addition, in order to consider not only the role of individual grass species in soil bacteriobiome modification but also to emphasize functional types of grasses, the fodder grasses: *Lolium perenne* (Lp), *Festuca arundinacea* (Fa)*,* and *Phleum pratense* (Pp) were grouped and marked as T1, whereas the lawn grasses: *Lolium perenne* L.× *hybridum* (Lh)*, Poa pratensis* (Pr), and *Festuca rubra* (Fr)—as T2.
