Fig. 1: Schematic illustration of the experimental design.

a Design of the microcosm experiment to manipulate the soil microbial community by subjecting soil suspensions to a range of removal treatments (three antibiotics, one antimicrobial peptide, four size filtration, one heat shock, and one oxidative stress), targeting various microbial groups (n = 25). Pie charts are symbolizing the composition of microbial communities. Values in parenthesis indicate the number of replicates. b Summary of possible ecological interactions between different species (A, B, and C) in the community (before) and consequences in each case for the other two species (increased or decreased relative fitness represented by the size of the symbols) when species A is depleted by a removal treatment (after). In examples 1–3, depletion of species A caused a decrease in relative fitness of species B, which it is not possible to distinguish from a decrease in relative fitness caused by a direct effect of the removal treatment. Only examples 4–6 causing an increase in the relative fitness are detected in our approach.