Abstract
Understanding the impact of microbial interactions on plants is critical for maintaining healthy native ecosystems and sustainable agricultural practices. Despite the reality that genetically distinct plants host multiple microbes-of-large effect in the field, it remains unclear the extent to which host genotypes modulate non-additive microbial interactions and how these interactions differ between benign/pathogenic environments. Our study fills this gap by performing a large-scale manipulative microbiome experiment across 7 genotypes of the model legume Medicago truncatula. We combine plant performance metrics, survival analyses, predictive modeling, RNA extractions, and targeted gene expression to assess how host genotype and microbes non-additively interact to shape plant growth and disease ecology. Our results reveal three important findings: (1) host genotypes with high tolerance to pathogens benefit more from multiple mutualist interactions than susceptible genotypes, (2) mutualists confer the same non-additive plant performance benefits in both benign and pathogenic environments, and (3) the quality of the symbiotic relationship with mutualists is a strong predictor of host survival against pathogenic disease. By applying these findings towards developing crops that promote synergistic microbial interactions, yields and pathogen defense could be simultaneously increased while reducing the need for toxic fertilizers and pesticides.