Predicting soil fungal communities from chemical and physical properties

Summary in the original language of the document

The information on the microbiome present in agricultural soil can be useful for making soil management decision. While it is well established that the diversity of soil microbes is largely controlled by environmental variables, microbiome community prediction from soil properties is emergent. This study examined whether soil physico-chemical properties could be used to predict the composition of soil fungal communities through multivariate ordination. Soil samples were collected from 59 arable fields in Switzerland, and physico-chemical soil properties were paired with profiles of soil fungal communities. The composition of fungal communities was characterized using long-read sequencing of the entire ribosomal internal transcribed spacer. Redundancy analysis was used to combine the physical and chemical soil measurements with the fungal community data. We identified a set of 10 soil properties that explained the fungal community composition. Soil properties with the strongest impact on the fungal community included pH, potassium and sand. The model was evaluated using leave-one-out validation and proved to be successful for most soils, with only 4/59 soils having poor correlation coefficients between observed and predicted communities. Prediction was less successful for soils with unique properties or diverging fungal communities, while it was most successful for soils with similar characteristics. Reliable prediction of microbial communities from chemical soil properties could eliminate the need for complex and laborious sequencing-based generation of microbiota data. This could make soil microbiome information available for agricultural purposes such as pathogen monitoring, field inoculation or yield projections.