Objective Soil microorganisms influence soil fertility by participating in the decomposition of organic residues, the mineralization of insoluble nutrients, and the transformation among different nutrient forms. This study aimed to delve into the microbial characteristics of grain fields with varying fertility levels and their intrinsic links to soil fertility in Huang-Huai-Hai region of China.

Methods Wuqiang County in Hebei Province is a typical winter wheat-summer maize producing county. Soil were sampled using a 3 km×3 km grid pattern across the entire county, resulting a total of 43 representative soil samples. These samples underwent testing for total N, organic matter, available P, available K, pH, and soil respiration rate to calculate soil fertility indexes (SFI) of the represented farmlands. Additionally, the extracellular enzyme activities were determined using specific kids, and the microbial communities were extracted and identified using high-throughput sequencing technology. The beneficial bacteria and pathogenic fungi were pointed using databases dedicated to plant beneficial bacteria and fungal pathogens. Correlation analysis and random forest prediction were employed and a structural equation model based on partial least squares path modeling was constructed to investigate the interrelations between microbial communities and soil fertility, along with its key indicators.

Results High-fertility (HF) soils exhibited higher total N (41%), organic matter (47%), available K (38%), and available P (82%) than the low fertility (LF) soils. The soil total N (P=0.03) and organic matter (P=0.04) significantly affected the composition of microorganisms in soils of different fertility levels. HF soils also demonstrated a significantly reduction in bacterial α-diversity and distinct species composition compared to LF soils. Specially, the relative abundance of the genus Nitrososphaera in HF soils were 48% and 63% higher than that of medium and low fertiltity soils, respectively. Similarly, the beneficial genus Methylophilus showed a 69% and 141% increase in relative abundance, while the pathogenic genus Exophiala decreased by 40% and 67%, respectively. HF soils also showed a 0.25% decrease in the bacterial Shannon index compared to LF soils, along with a 39% and 37% increase in the activities of alkaline phosphatase and glucosidase respectively. Random forest prediction indicated the significant impact of genera Nitrososphaera, Methylophilus, and Exophiala on soil fertility (P<0.01). The structural equation model showed that microbial composition had an extremely significant impact on microbial diversity ( P<0.01), and by enhancing soil extracellular enzyme activity, the nutrient status of the soil was significantly enhanced (P<0.05).

Conclusion The composition of microbial communities in soils with varying fertility levels shows significant differences. Specifically, the abundance of the ammonia-oxidizing archaea genus Nitrososphaera and the beneficial bacterium genus Methylophilus increases, while the abundance of the potentially pathogenic fungus genus Exophiala decreases. This optimized community structure enhances the activities of soil extracellular enzymes associated with carbon, nitrogen, and phosphorus cycling, facilitates the transformation of organic nutrients, and thereby effectively improves soil functioning. Given these findings, the abundance of differential microorganisms can be incorporated into soil fertility evaluation systems to provide a more comprehensive assessment of soil ecological functions.