Identifying genomic loci under selection in a widespread Bradyrhizobium, across Australian ecosystems using landscape genomics

Jose Antonio Blanco Alcantara, Anna Simonsen

Climate change is reshaping soil environments, intensifying selective pressures on microbial communities that drive essential ecosystem processes. Understanding how nitrogen-fixing rhizobia adapt to environmental variation is critical for predicting ecosystem responses to global change. Here, we used redundancy analysis (RDA) to identify genomic loci associated with environmental gradients across 375 Bradyrhizobium diazoefficiens isolates from southwestern Australia, analyzing variants mapped to four independent reference genomes.
Climate variability emerged as the dominant driver of genomic adaptation, with annual rainfall and temperature ranking first and second among twelve environmental variables tested. These climate effects substantially exceeded those of soil chemistry factors, with rainfall explaining nearly twice the variance of the strongest soil variable (carbon content). Notably, annual temperature showed its strongest association with RDA axis 3 rather than early axes, demonstrating that limiting analysis to the first two RDA dimensions can cause researchers to miss important environmental effects.
Functional enrichment analysis revealed that signal transduction mechanisms were significantly over-represented among outlier loci, while core symbiosis genes showed consistent depletion, providing suggestive evidence for purifying selection maintaining nitrogen-fixing capacity. Within symbiosis genes, regulatory components (exoR, regB) showed 100% outlier rates while structural machinery (T4SS, nol genes) showed complete conservation, supporting a “regulatory evolution” model where adaptation occurs through expression control rather than structural changes.
These findings demonstrate that Bradyrhizobium populations adapt to climate heterogeneity primarily through signal transduction and regulatory networks, while core metabolic and symbiotic functions remain under strong functional constraint.