Objective This study aimed to investigate the changes in wheat rhizosphere fungal community structure and functional guilds under different fertilization treatments. By quantifying the direct and indirect contributions of soil chemical properties and the rhizosphere fungal community to wheat theoretical yield, this research provides a theoretical basis for optimizing fertilization management to improve the rhizosphere micro-ecological environment.

Methods A long-term localized fertilization experiment was established in 2005 in a fluvo-aquic soil region of Shijiazhuang, North China Plain. The experiment included eight treatments: no fertilizer (CK), application of a single chemical fertilizer (N, P, K), imbalanced fertilization (NP, NK, PK), and balanced fertilization (NPK). Soil chemical analysis, high-throughput sequencing, and structural equation modeling (SEM) were comprehensively employed to systematically investigate the effects of different fertilization treatments on soil nutrients, rhizosphere fungal community structure, and wheat theoretical yield.

Results No significant differences in soil ammonium nitrogen (NH₄⁺-N) content were observed among treatments. However, nitrate nitrogen (NO₃^--N) content significantly accumulated in nitrogen-applied treatments (N, NK, NP, NPK). Soil available phosphorus (Olsen-P) content was significantly increased in phosphorus-applied treatments (P, NP, PK, NPK). Soil available potassium content was significantly higher in potassium-applied treatments (K, NK, PK, NPK) compared to non-potassium treatments. The soil pH in the NPK treatment was significantly lower than that in all other treatments. Regarding soil organic matter (SOM) and total nitrogen (TN), all fertilization treatments were significantly higher than the non-fertilized control (CK). Specifically, compared to CK, SOM content was significantly increased by 47.8% and 30.4% in the NP and NPK treatments, respectively; while TN content was significantly increased by 43.6% and 30.9% in the NP and NPK treatments, respectively. Fertilization significantly affected wheat theoretical yield, with the NPK treatment achieving the highest theoretical yield, followed by the NP treatment. All fertilization treatments significantly altered the rhizosphere fungal community structure, and the rhizosphere soil exhibited a more sensitive response to fertilization than the bulk soil. The NPK treatment reduced the relative abundance of potential pathogens, such as Alternaria and Phoma, while significantly increasing the abundance of the beneficial saprophytic fungus Mortierella in the rhizosphere soil. Soil chemical properties especially available phosphorus, organic matter, and total nitrogen had a direct positive effect on theoretical yield. In contrast, the rhizosphere fungal community exerted an indirect effect on theoretical yield through complex interaction networks, particularly the trade-off between saprophytic and pathogenic fungi. Consequently, the NPK treatment achieved the highest theoretical yield and the lowest relative abundance of pathogens in the fungal community, thereby attaining the dual objectives of "yield increase" and "pathogen suppression".

Conclusion In the typical fluvo-aquic soil of north China, balanced fertilization with NP or NPK is more effective in enhancing key soil fertility indicators—organic matter and total nitrogen content—than applying a single chemical fertilizer or imbalanced treatments like NK and PK. Balanced NPK application directly increases wheat yield by elevating critical fertility levels, including organic matter, total nitrogen, available phosphorus, and available potassium. Furthermore, NPK fertilization significantly promotes the growth of beneficial functional fungal groups in the rhizosphere, particularly saprotrophic fungi (e.g., Mortierella), while suppressing the relative abundance of potential pathogenic fungi (e.g., Alternaria and Phoma), thereby indirectly boosting wheat yield through microbial pathways.