Seasonal population structure and adaptive signatures across long-term marine microbial time series

Long-term microbial population dynamics can reveal how ocean microbiomes respond to environmental variation, but resolving within-species change requires suitable genomic references. We integrated 30 PacBio HiFi long-read metagenome-assembled genomes (MAGs) from the northwestern Mediterranean Sea with monthly Illumina metagenomes from neighboring time series spanning 12 and 7 years. Population differentiation varied widely among genomes and often differed in magnitude between sites, while a subset showed increasing differentiation with temporal distance. Genetic diversity and population structure were frequently seasonal, with populations associated with warm or cold waters and others spanning both thermal regimes. Candidate adaptive genes accounted for 1.6-3.0% of coding sequences, and many were detected in both time series. Their pN/pS trajectories included regular seasonal fluctuations consistent with changing strain or ecotype contributions, alongside irregular temporal patterns. These results show that combining long-read MAG references with dense short-read time series can resolve seasonal population structure, site-specific differentiation, and candidate adaptive variation within marine microbial species, providing a framework for tracking contemporary population-genomic change in the ocean.