Fig. 1

Schematic diagram for experimental design, rhizosphere sample collection and following validation experiments. a A semi-natural rhizobox sampling system was embedded in natural tomato field soil to repeatedly sample rhizosphere soil for each plant by removing individual nylon mesh bags inserted in the ‘middle’ layer of the rhizobox, which were immediately stored at − 80 °C. b Bacterial wilt disease development was quantified through time within one field plot where red and blue circles denote diseased and healthy plants, respectively. c Sample processing flow where four healthy and diseased plants were selected for further temporal analysis and their respective samples from earlier time points were sequenced using shotgun metagenomics. d Validation pipeline where bacterial and phage species were isolated from the same field and their effects on the pathogen and bacterial wilt disease outcomes measured in vitro and in planta. All panels were created with BioRender.com