Abstract
BACKGROUND: Healthy soil is crucial for maintaining the sustainability of soil ecosystem services and ensuring food security. The soil microbiome serves as a key indicator of soil health. However, long-term continuous monoculture significantly adversely affects the diversity and functioning of soil microbial communities, posing a serious threat to sustainable agricultural development. Therefore, making it crucial to understand the microbial mechanisms underlying these challenges. METHODS: In this study, tobacco was subjected to different planting durations: 1 year (CR), 5 years (CC5), 10 years (CC10), and 15 years (CC15). The rhizosphere microbial community assembly process, composition, keystone taxa, and their relationship to continuous cropping challenges were analyzed. RESULTS: The rhizosphere bacterial community structure of tobacco after 5 years of continuous cropping was significantly separated from other treatments, while no significant separation was observed in the fungal community. Further investigation into the assembly processes of microbial communities under different continuous cropping durations revealed that bacterial community assembly processes exhibited differences, whereas no significant differences were observed in fungal community assembly processes. Specifically, the rhizosphere bacterial community in CC5 was predominantly shaped by deterministic assembly processes, explaining its structural distinctiveness from other treatments. Co-occurrence network analysis revealed that the number of nodes and edges in bacterial-fungal interactions decreased by 22.70% and 79.86%, respectively, in CC5 compared with CR. Differential microbial abundance identified a significant decline in key microbes (Rhodanobacter, Ellin6067, Frankiales, and Setophoma) alongside a marked increase in the abundance of genus Verticillium in CC5 relative to CR. RDA indicated these bacterial genera correlated negatively with pathogen accumulation and disease incidence but postively with yield These relationships potentially constituted the primary driver for exacerbated continuous cropping obstacles in CC5. In CC10, bacterial community assembly was primarily dominated by stochastic processes. Although the number of network nodes and edges increased by 21.96% and 204.73%, respectively, compared to CC5, they remained lower than those in CR. By extending the continuous cropping to 15 years, the bacterial community assembly was shaped by stochastic processes. Compared with CC5, the number of network nodes and edges in CC15 increased by 34.39% and 405.44%, respectively, with beneficial microbes (Rhodanobacter, Ellin6067, Frankiales, and Setophoma) showing remarkable abundance recovery and pathogenic fungi like Verticillium declining, ultimately resulting in microbial community characteristics approximating those observed in the CR. CONCLUSIONS: The bacterial community in CC5 was driven by deterministic assembly processes, resulting in a distinct structure. The complexity and stability of the co-occurrence network significantly decreased, accompanied by a higher abundance of harmful microbes and a lower abundance of beneficial microbes, which exacerbated the continuous cropping obstacles. In contrast, after 15 years of continuous cropping, the bacterial community assembly shifted to stochastic processes and trans-kingdom co-occurrence network complexity and stability strengthened, beneficial microbes increased, and continuous cropping challenges alleviated.