Global geography outweighs long-term dynamics in shaping marine microbial population structure

Abstract

Marine microbial populations play essential roles in ocean ecosystems, yet the processes shaping their genomic structure across space and time remain poorly understood. Here, we examined the population-scale patterns of 1,505 prokaryotic metagenome-assembled genomes (MAGs) retrieved from the Northwestern Mediterranean Sea in two long-term coastal time series, along 15 and 7 years in these sites, as well as in the global ocean. We found that populations were generally more genomically differentiated across large spatial scales than across long temporal scales. Among a subset of 389 MAGs well represented in all datasets, 68.4% showed weak population divergence over time but strong global-scale differentiation. This was evident in the abundant and widespread cyanobacteria Prochlorococcus and Synechococcus. In contrast, only 6.4% of the MAGs exhibited weak divergence over time and space, with SAR11 MAGs being a clear example, likely reflecting their high dispersal and recombination rates. Other groups, such as SAR86 and Flavobacteriales, showed strong divergence at temporal and spatial scales, suggesting seasonal and/or regional adaptation. Positive selection was more readily detectable in the long-term coastal observatories than in the global ocean, despite the more significant population divergence observed across broad geographic scales. Temperature consistently showed a significant association with the population structure of many MAGs. Overall, our results highlight the dominant influence of large geographic scales in shaping microbial population structure alongside taxon-specific responses to temporal variation, particularly seasonality. Altogether, our work advances the understanding of microbial population structure across broad spatial and temporal scales, a critical step toward predicting microbial dynamics in a changing ocean.