F GSCHWEND¹, F WIDMER<sup>1

1</sup>Molecular Ecology, Institute for Sustainability Sciences, Agroscope, 8046 Zürich, Switzerland

Soil samples (N=152) originating from 9 different Swiss cantonal soil monitoring surveys were analysed using an amplicon-based metabarcoding approach, i.e. PCR-based amplification and high-throughput sequencing of genomic markers. We assessed alpha and beta diversity of soil bacterial and fungal communities and statistically tested the contribution of several environmental factors to their distinction. DNA extraction yielded concentrations from 7 to 198 µg/g dry soil. Bacterial and fungal diversities were detected by amplicon sequencing of the ribosomal SSU (V3-V4) and ITS2 regions, respectively. After thorough sequence filtering, we obtained at least 4078 bacterial and 9725 fungal sequences per sample. Clustering operational taxonomic units (OTU) at 97% sequence identity resulted in a total of 6944 fungal and 11258 bacterial OTU, corresponding to 317-472 and 881-1349 OTU per sample. Arable land harboured less fungal, but greater bacterial richness in comparison to grassland sites. Carbon content, pH, and other environmental variables revealed only weak correlations to fungal and bacterial alpha diversity. However, beta diversities, i.e. diversity of microbial community structures among sites, revealed strong responses to different factors. DISTLM analysis indicated that pH, land-use, clay content, and altitude were the most important forces shaping both bacterial and fungal community structures in our dataset. With 22.7% of total variance, pH was the strongest factor for bacterial communities, while fungal communities lacked such a strongly influencing factor. For them, land-use was the most important factor explaining 5.2% of the total variance. A canonical analysis of principal coordinates (CAP) showed that arable land and grassland harboured distinctly structured microbial communities. Reclassification of single samples to the two land-use types yielded 96.7% and 96.1% successful classifications for bacterial and fungal communities, respectively. These results build a step towards the establishment of a baseline for classifying soil microbial diversity in Swiss land-use types and the development of a soil biomonitoring system based on amplicon sequencing.