Dynamic succession patterns, nitrogen cycling potential, and multi-scale assembly mechanisms of cross-habitat bacterial communities in lakes driven by seasonal frozen conditions

Microorganisms are key bioindicators of aquatic environment, yet their dynamics under seasonal ice cover-affecting 50 % of global lakes-remain poorly understood. This study comprehensively employed metagenomics and bioinformatics to analyze the diversity characteristics, species composition, nitrogen cycling potential, and community assembly mechanisms of bacterial communities during frozen and non-frozen periods. Results showed that bacterial species richness and diversity in water were significantly higher during the frozen period compared to the non-frozen period, with both metrics consistently higher in water than in sediment. In winter, ice formation significantly reshaped the bacterial community structure in water, while exerting no notable disturbance on the sediment bacterial community composition. Freezing exerts contrasting regulatory effects on the primary nitrogen cycling functions of bacterial communities in the water column versus the sediments. In the water column, the potential of ammonia assimilation is significantly suppressed during freezing, whereas mineralization and assimilatory nitrate reduction to ammonium persist. In contrast, within the sediments, ice cover generally enhances the activity of major nitrogen transformation pathways, including ammonia assimilation, mineralization, and nitrification. Notably, sediment nitrogen fixation potential is nearly four times higher in non-frozen periods compared to frozen periods. Stochastic processes dominate bacterial community assembly, while the freezing process shifts the dominant role from drift to dispersal limitation. However, in deterministic processes, heterogeneous selection serves as a key regulatory factor. The study revealed the adaptive strategies of bacterial communities to freezing in shallow lakes of cold-arid regions, providing a theoretical basis for ecological risk prediction in frozen lakes and ecological management of shallow lakes in cold-arid regions.