Strong pairwise interactions do not drive interactions in a plant leaf associated microbial community

Microbial communities that promote plant growth show promise in reducing the impacts of climate change on plant health and productivity. Understanding microbe-microbe interactions in a community context is paramount for designing effective microbial consortia that enhance plant resilience. In this study, we investigated the dynamics of a synthetic microbial community (SynCom) assembled from <i>Arabidopsis thaliana</i> leaves to elucidate factors shaping community composition and stability. We found notable disparities between <i>in vitro</i> pairwise interactions and those inferred from correlation networks <i>in planta</i>. Our findings suggested that secondary metabolites, particularly antimicrobials, might mediate interactions <i>in vitro</i>, but are not key drivers of microbial interactions in a community context. Through co-cultivation experiments, we identified the siderophore pseudobactin as a potent antimicrobial agent against several SynCom members, but its impact on community composition <i>in planta</i> was negligible. Notably, dominant SynCom members, such as <i>Pseudomonas koreensis, Flavobacterium pectinovorum</i>, and <i>Sporobolomyces roseus</i>, exhibited only positive correlations, suggesting synergism based on factors such as exopolysaccharides and biotransformation might drive community dynamics rather than competition. Two correlations between SynCom members in the co-abundance network corresponded to their pairwise <i>in vitro</i> interactions, highlighting the potential for further research and demonstrating the usefulness of correlation networks in identifying key microbe-microbe interactions. Our findings highlight the importance of considering microbiome-wide interaction studies and synthetic communities in understanding and manipulating plant microbiomes.