Towards standardized experimentation in soil research – a synthetic ecology approach

Soils contain a vast diversity of microorganisms including millions of cells and thousands of species. Many of those species encode for similar functions which is known as functional redundancy. From an ecosystem perspective, it remains unknown how many of such species are contributing to processes and are actually necessary to perform functions. Not knowing about the number and type of taxa of a soil sample a priori, let alone of the interaction between those taxa or the chemical environment of their habitats, hampers the reproducibility of soil ecological research and renders a targeted experimentation virtually impossible. This is one of the most important reasons why linking microbial identity to processes at the soil scale has proven to be everything but easy.

A promising avenue to overcome these limitations are model ecosystems that allow to identify principles of community functioning via standardised experimentation. Here, we will present a synthetic ecology approach using artificial soil mimicking the structural and chemical complexity of a soil and a synthetic microbial community to investigate microbial functioning in soil. Our approach includes genome-guided in silico design of synthetic communities to reproduce the functionality of soil heterotrophic bacteria and fungi, while largely decreasing the number of individual taxa for laboratory experimentation.

Initial experiments suggested that our synthetic community was able to establish in artificial soil and sterilized natural soil and to perform a range of simple soil processes, as evident through potential enzymatic activities, heterotrophic respiration, and changes in organic and inorganic soil components. Molecular analysis of the community composition over time demonstrated high similarities of the established community among replicates, indicating low effects of stochasticity on community assembly, a major requirement for reproducible experimentation. These promising preliminary data indicate that our model system could indeed represent the experimental platform for targeted experimentation in soil ecological research in the future.