Objective This study investigated the changes in rhizosphere soil microbial community composition and network interactions between healthy and diseased Pogostemon cablin plants and elucidated the relationship between soil physicochemical factors and rhizosphere microbial community variation.

Method Healthy and diseased P. cablin plants were selected. Soil physicochemical properties were analyzed, and high-throughput sequencing of 16S rRNA and ITS was performed to systematically characterize the changes in the composition, structure, functional potential, and microbial interactions of rhizosphere bacterial and fungal communities during disease occurrence.

Result Soil pH, organic matter, and hydrolyzable nitrogen were the most significant environmental factors driving variations in both bacterial and fungal communities. A total of 12 463 bacterial ASVs and 3 468 fungal ASVs were obtained from the rhizosphere soil of healthy and diseased P. cablin plants. The proportion of unique ASVs was lower in the diseased group, and marked differences in community composition were observed between the two groups. The relative abundance of Proteobacteria in the healthy group was 30.00%, higher than that in the diseased group (27.44%), whereas the relative abundance of Actinobacteria in the diseased group reached 26.21%, higher than that in the healthy group (21.53%). At the genus level, the healthy group was significantly enriched with potentially beneficial bacteria such as Bradyrhizobium, Ramlibacter, and Novosphingobium, as well as potential biocontrol fungi such as Arthrobotrys and Trichoderma. The diseased group was significantly enriched with taxa such as Streptomyces, Pseudarthrobacter, Microbacterium, Acremonium, and Phialemonium. However, the relative abundances of Ralstonia and Fusarium did not differ significantly between the two groups. Network interaction analysis showed that the edge number and average degree of the bacterial network in the diseased group were 780 and 15.60, respectively, significantly lower than those in the healthy group (1 626 and 33.53), and the modularity coefficient increased from 0.19 to 0.32, indicating a looser and more fragmented network structure. In contrast, the edge number and average degree of the fungal network in the diseased group were 1 421 and 28.71, respectively, markedly higher than those in the healthy group (1 142 and 23.31), indicating a more tightly aggregated network. Functional prediction indicated that nitrogen fixation was significantly enriched in the healthy group, whereas the relative abundances of nitrate reduction, nitrate respiration, and hydrocarbon degradation were significantly increased in the diseased group. However, fungal trophic mode composition did not differ significantly between the two groups.

Conclusion The occurrence of soil-borne disease in P. cablin was accompanied by coordinated changes in rhizosphere microbial community composition, interaction networks, and ecological functions, with bacterial and fungal communities exhibiting distinct response patterns. This study reveals the potential mechanisms of rhizosphere microecological homeostasis shifts during soil-borne disease development and provides a theoretical basis for ecological disease regulation and microbial resource utilization.