The experimental site was located in a wheat farm in Ningjin County, Xingtai City, Hebei Province, China (37° 37’ N, 114° 53′ E, 18 m a.s.l.), which has a subtropical monsoon climate. The area has an average annual temperature of 12.5°C and an average annual rainfall of 1,150–1,550 mm. Winter wheat is rotated with summer maize. The winter wheat variety used was Jimai 22, and the maize variety was Xianyu 335. In 2022, during the wheat filling stage (mid-May), three plots (200 m²) with different severities of wheat crown rot—light (L), moderate (M), and severe (S) were selected from the wheat farm. Following the “W” multiple-point sampling method, wheat rhizosphere soil was collected. For each plot, soil samples were collected in three replicates, with each replicate including 10 sampling points, and 20 wheat plants per sampling point. Using a sampling spade, the entire wheat root system was carefully excavated, loose soil was shaken off and collected for greenhouse experiment, then whole wheat roots were subsequently placed in sterilized PBS buffer and shaked for 30 min, finally the rhizosphere soil was collected by centrifugation at 12,000 g for 5 min (Wu et al., 2024). The remaining wheat plants were used to investigate the incidence of wheat crown rot. A portion of the soil sample was stored at −80°C for DNA extraction, and the remainder was stored at 4°C for analyses of soil chemical properties.

Disease severity of WCR at the filling stage was rated on a 0–4 scale based on the symptoms observed on the crown (Zhang et al., 2022). 0: completely healthy; 1: light browning on the crown, less than 25% necrosis; 2: 25–50% necrosis; 3: 51–75% necrosis; 4: greater than 75% necrosis, or show symptoms of dried white ears or complete plant death. A disease index was then calculated as ∑ (Number of diseased plants with each score × Highest score)/(Total number of plants × Highest score) × 100%.

Genomic DNA from 0.5 g rhizosphere soil was extracted using the E.Z.N.A. Soil DNA Kit (Omega Bio-TEK, Norcross, GA, United States) following the protocol provided by the manufacturer. DNA concentration was assessed with a TBS-380 Mini-Fluorometer (Turner Biosystems, CA, United States) and its purity was assessed with NanoDrop 2000 UV–Vis spectrophotometer (Thermo Scientific, Wilmington, DE, United States). DNA quality was confirmed using 1.2% agarose gel electrophoresis. To assess the titers of F. pseudograminearum in the rhizosphere soil, qPCR was conducted using primers Fptri3eF (5’-CAAGTTTGATCCAGGGTAATCC-3′) and Fptri3eR (5’-GCTGTTTCTCTTAGTCTTCCTCA-3′) as described by Obanor and Chakraborty (2014). A 20 μL reaction mixture was prepared using 1 μL of template DNA, following the protocol provided by TaKaRa Ex Taq HS DNA Polymerase. Amplification conditions included an initial denaturation at 95°C for 30 s, followed by 40 cycles of 95°C for 5 s and 60°C for 34 s, performed on an ABI7500 Real-Time PCR system. Each reaction was run in triplicate. Genome copies of F. pseudograminearum were quantified based on a standard curve generated from a recombinant plasmid containing the target DNA sequence.

Rhizosphere soil samples were dried, ground and sieved before soil analysis. Soil chemical property was measured with the methods of Wang et al. (2021). Soil pH was measured using the potentiometric method with a soil-to-water ratio of 1:5 (mass-to-volume ratio). Soil organic carbon was determined using the potassium dichromate oxidation method with dilution and heating. Available phosphorus was measured using the acid-soluble molybdenum-antimony colorimetric method, and available potassium was determined by the ammonium acetate extraction method with flame photometry.

The bacterial V3–V4 region of the 16S rDNA was amplified using specific primers 338F (5’-ACTCCTACGGGAGGCAGCAG-3′) and 806R (5’-GGACTACHVGGGTWTCTAA T-3′), and the fungal internal transcribed spacer-1 (ITS-1) region was amplified using specific primers primers ITS1F (5’-CTTGGTCATTTAGAGGAAGTAA--3′) and ITS2R (5’-GCTGCGTTCTTCATCGATGC--3′). The PCR amplification system was set up in a 20 μL reaction volume: 4 μL of 5 × FastPfu buffer, 2 μL of 2.5 mmol·L⁻¹ dNTPs, 0.8 μL of each primer (5 μmol/L), 0.4 μL of FastPfu polymerase, and 1 μL of DNA template (20 ng/μL), with ddH₂O added to a final volume of 20 μL. The PCR cycling conditions were as follows: an initial denaturation at 95°C for 3 min, followed by 27 cycles of denaturation at 95°C for 30 s, annealing at 55°C for 30 s, and extension at 72°C for 30 s, with a final extension at 72°C for 10 min. PCR products were checked by 1.5% agarose gel electrophoresis, recovered, and purified. Samples were then sent to Shanghai Majorbio Bio-Pharm Technology Co., Ltd. for library construction and high-throughput sequencing on the Illumina MiSeq platform (Illumina, San Diego, United States) according to standard protocols.

Raw sequencing data were quality-controlled using Trimmomatic software, and sequences were assembled with FLASH software (Magoc and Salzberg, 2011). (1) A 50 bp sliding window was applied, and bases in the trailing sequence were trimmed when the average quality score within the window was below 20. Sequences shorter than 50 bp after quality control were removed. (2) Paired-end reads were merged based on overlapping regions with a minimum overlap length of 10 bp and a maximum mismatch rate of 0.2; unmergeable reads were discarded. (3) Sequences were demultiplexed based on barcodes and primers, with sequences containing ambiguous bases being removed. This process produced high-quality sequences for subsequent analysis. The raw sequences used in this study have been deposited in the NCBI (National Center for Biotechnology Information) database under the Sequence Read Archive (SRA) accession number PRJNA881807.

A total of 1 g of rhizosphere soil with slight occurrence of wheat crown rot was subjected to serial dilution and plated on LB agar (tryptone 10 g, yeast extract 5 g, NaCl 10 g, agar 15 g, ddH₂O up to 1 L, pH 7.0), followed by incubation at 37°C. Bacterial colonies with different colony characteristics were isolated and maintained on LB agar for further analysis. To assess the inhibition of F. pseudograminearum mycelial growth, a 5 mm agar plug of the fungus was placed in the center of a PDA plate (potato infusion 200 g, glucose 20 g, agar 15 g, ddH₂O up to 1 L), bacterial isolates (ddH₂O was used as control) inoculated 2.5 cm away from the fungal plug. After incubation at 28°C for 5 days, the inhibition zone between fungal and bacterial colonies was measured. The reaction to Gram staining were tested (Li et al., 2021). Genomic DNA of BF-237 was extracted using the MiniBEST Bacterial Genomic DNA Extraction Kit Ver. 3.0 (Takara, Beijing, China). The gene of gyrB was amplified with primers UP-1S and UP-2Sr (Dong et al., 2023) and sequenced (Tsingke Biotechnology, Beijing, China). The resulting gyrB sequence was analyzed using blastn against the NCBI nr database to identify similarities. Additionally, 12 other gyrB sequences were selected for phylogenetic analysis.

In a greenhouse experiment, the biocontrol effect of BF-237 on wheat crown rot in sterilized field soil and in untreated field soil were investigated. Bulk soil with severe wheat crown rot was collected as described in method 2.1. Half of the collected soil was directly used for greenhouse experiment, the remaining half of the soil was autoclaved before use. The soil was autoclaved using the liquid cycle for 30 min at 121°C, then cooled at room temperature, then we repeated this procedure, sterilized soil was stored at 4°C until use. Soak wheat seeds in 2% sodium hypochlorite for 5 min, then rinse three times with water. Surface sterilized wheat seeds (Shixin 828) were soaked in 10⁹ bacterial cells/mL BF-237 for 4 h at 28°C. F. pseudograminearum was grown in the sodium carboxymethyl cellulose medium (CMC, carboxymethyl cellulose 10 g, KH₂PO4 1 g, MgSO₄·7H₂O 0.5 g, FeSO₄·7H₂O 0.01 g, NaCl 0.5 g, yeast extract 1 g, ddH₂O up to 1 L) broth at 28°C and 180 rpm for 5 days to produce spores. Five treatments were set up, (1): SCK1, sterilized seeds were planted in sterile soil; (2): SCK2, sterilized seeds were planted in sterile soil with F. pseudograminearum (10⁵ spores/g soil); (3): SB237, seeds soaked with BF-237 were planted in sterile soil with F. pseudograminearum (10⁵ spores/g soil); (4): FCK, sterilized seeds were planted in field soil; and (5): FB237, seeds soaked with BF-237 were planted in field soil. There were 4 replicates per treatment, 5 pots per replicate, and 20 seeds per pot. The pots were placed in a greenhouse with 16 h light/8 h dark at 25°C. At 21 days after sowing, wheat height was measured, and the severity and severity index of wheat crown rot were classified and calculated as described in method 2.2.

The alpha diversity (Shannon and Simpson indices) of the bacterial and fungal community were calculated by Mothur v.1.30 ¹ and the vegan package in R. Beta diversity was examined by Principal Coordinate Analysis (PCoA) based on Bray-Curtis distances in R. Redundancy analysis (RDA) was executed in R to analyze the relationship between dominant taxa of microorganisms and soil properties. Correlation networks were constructed to investigate the relationships among environmental factors, bacterial diversity indices, and wheat crown rot, employing the igraph and psych packages in R. All data underwent normality checks with the Shapiro–Wilk test prior to statistical analysis in SPSS v.16.0 (SPSS, Chicago, IL, United States). For non-normally distributed data, including alpha diversity indices, wheat crown rot, and the relative abundance of key genera, we applied the Kruskal-Wallis test for significance, using SPSS v.16.0.

According to field survey results, the disease indices of wheat crown rot in light (L), moderate (M), and severe (S) plots were 2.4, 14.2, and 41.6, respectively (Figure 1A). The quantity of F. pseudograminearum in the rhizosphere soil for the L, M, and S plots was 5.58 × 10³ copies/g, 2.25 × 10⁴ copies/g, and 3.26 × 10⁵ copies/g (Figure 1B), indicating a positive correlation between the incidence of wheat crown rot and the concentration of F. pseudograminearum in the rhizosphere soil. Following the occurrence of wheat crown rot, the contents of organic matter, available phosphorus, and available potassium in the wheat rhizosphere soil showed a decreasing trend. Compared to the L, these three indicators in the M rhizosphere soil decreased by 0.37, 8.18, and 23.77%, respectively. While in S, the reductions reached 2.59, 29.84, and 24.63%, respectively. Additionally, soil pH gradually decreased with the progression of wheat crown rot. The pH in the S plot was 7.46, which significantly lower than in the L pot (7.88) (Table 1).

In this study, sequencing data were rarefied to the minimum sample size for alpha diversity index analysis. Results showed that coverage rates for all treatments exceeded 98%, indicating that the bacterial and fungal sequences obtained in this study achieved good coverage and that the sequencing depth was sufficient for analyzing bacterial and fungal diversity. The number of bacterial and fungal OTUs decreased as the disease index of wheat crown rot increased. The Sob, Chao, and Shannon indices for bacteria showed no significant change between L and M levels but were significantly lower in S, which increased by 11.61, 7.10, and 14.34%, respectively. For fungi, the Sob, Chao, and Shannon indices decreased with increasing disease index; compared to L, these indices in M decreased by 13.99, 10.72, and 10.62%, respectively, and in S by 27.97, 20.87, and 24.78%, respectively (Table 2). These results indicate that the occurrence of wheat crown rot reduced the alpha diversity of bacteria and fungi in the wheat rhizosphere soil.

The effect of wheat crown rot on the beta diversity of bacterial and fungal communities in the wheat rhizosphere soil was analyzed using PCoA. For bacteria (Figure 2A), the PCoA results indicated that the PC1 and PC2 axes explained 50.18 and 11.65% of the variance in the bacterial community, respectively. Between L and M, there was little difference in bacterial community structure on the PC1 axis, with a 0.16 difference on the PC2 axis. However, in S, the bacterial community structure showed a significant difference on the PC1 axis compared to L and M, with a variance of approximately 0.3 (Supplementary Table S1). For fungi (Figure 2B), the PCoA results showed that the PC1 and PC2 axes explained 55.31 and 33.25% of the variance, respectively. In L and M, fungal community structure had a minor difference of 0.18 on the PC1 axis and a difference of 0.37 on the PC2 axis. In contrast, in S, fungal community structure showed significant differences on the PC1 axis compared to L and M, with differences of 0.32 and 0.50, respectively (Supplementary Table S1). These results demonstrate that wheat crown rot significantly altered the bacterial and fungal community structure in wheat rhizosphere soil on the PC1 axis.

Phylum-level analysis of bacterial community composition in wheat rhizosphere soil revealed 10 dominant phyla with relative abundance >1%, including Actinobacteria, Proteobacteria, Acidobacteria, Chloroflexi, Bacteroidetes, Gemmatimonadetes, Saccharibacteria, Firmicutes, Nitrospirae, and Verrucomicrobia (Figure 3A). In L, M, and S, the relative abundances of Actinobacteria (33.30, 33.56, 40.67%), Chloroflexi (9.21, 9.97, 11.85%), Saccharibacteria (1.71, 1.94, 1.95%), and Verrucomicrobia (0.96, 1.03, 1.09%) increased with the rise in wheat crown rot disease index. In contrast, the relative abundances of Acidobacteria (12.56, 11.80, 11.14%), Bacteroidetes (6.66, 6.37, 5.25%), and Nitrospirae (2.17, 1.45, 0.91%) decreased with the rise in wheat crown rot disease index. At the genus level (Figure 3B), grouping bacteria with <1% abundance and unclassified groups as “Other,” we found that the dominant genera with relative abundance >1% displayed varied trends. In L, M, and S, Bacillus (1.26, 1.34, 0.87%), Streptomyces (2.12, 1.74, 1.56%), Nitrospira (2.17, 1.45, 0.91%), Nonomuraea (1.21, 1.65, 0.20%), and Gaiella (1.45, 0.90, 0.76%) decreased with the rise in wheat crown rot disease index. However, Blastococcus, Sphingomonas, Skermanella, Pedobacter, Roseiflexus, Massilia, and Rubrobacter increased with the rise in wheat crown rot disease index.

At the phylum level (Figure 3C), the composition of the wheat rhizosphere fungal community was analyzed, and the results showed that the dominant phyla were Ascomycota, Zygomycota, and Basidiomycota. Their relative abundances ranged from 89.18 to 95.20%, 3.17 to 9.27%, and 0.32 to 1.08%, respectively. Compared to L, there were no significant differences in the relative abundance of these three phyla in M. However, in S, the relative abundance of Ascomycota decreased by 6.31%, while the relative abundance of Zygomycota and Basidiomycota increased by 1.92 and 2.38 times, respectively. At the genus level (Figure 3D), fungal groups with a relative abundance of less than 1% and unclassified groups were categorized as “others.” In L, M, and S, the relative abundances of Alternaria (12.47, 13.15, 27.55%), Rhizopus (3.32, 3.02, 8.82%), Oedocephalum (0.001, 1.40, 5.51%), Gibberella (4.19, 5.28, 5.44%), and Cyphellophora (0.2, 0.49, 1.04%) increased with the rise in wheat crown rot disease index. In contrast, the relative abundance of Mycosphaerella (31.65, 21.13, 21.55%), Pyrenochaetopsis (8.18, 6.34, 5.23%), Acremonium (4.50, 4.86, 3.04%), Myrothecium (2.48, 1.05, 0.96%), Myrmecridium (1.86, 0.41, 0.24%), and Preussia (3.06, 0.11, 0.05%) decreased with the rise in wheat crown rot disease index.

The analysis of the relationship between environmental factors and genus-level bacterial changes in the rhizosphere soil showed that the first principal component (RDA1) and the second principal component (RDA2) explained 49.35 and 15.49% of all variables, respectively (Figure 4A). The bacterial community structure in the rhizosphere soil of wheat was partially overlapping between fields with light and moderate wheat crown rot, indicating similarity in community structure. However, in fields with severe wheat crown rot, the bacterial community structure in the rhizosphere soil was located further away on the RDA2 axis compared to fields with light and moderate wheat crown rot, with no overlap, suggesting a significant change in bacterial community structure in heavily affected fields. Additionally, RDA revealed that different environmental variables had varying impacts on the overall bacterial community. Available phosphorus (r² = 0.937; p = 0.003) had a highly significant effect on soil bacterial community composition (p < 0.01), while wheat crown rot (r² = 0.820; p = 0.017) had a significant effect (p < 0.05) (Supplementary Table S2). VPA analysis showed that wheat crown rot occurrence and changes in soil chemical properties explained 64.36% of the variation in soil bacterial diversity, with wheat crown rot accounting for 54.76%, soil chemical properties accounting for 54.36%, and their interaction accounting for 44.76%. Among soil chemical properties, available phosphorus had the highest explanatory power, at 45.81% (Figure 4C).

The analysis of the relationship between environmental factors and genus-level fungal changes in the rhizosphere soil indicated that the first principal component (RDA1) and the second principal component (RDA2) explained 48.74 and 35.38% of all variables, respectively (Figure 4B). The fungal community structure in the rhizosphere soil of wheat showed no overlap among fields with light, moderate, and severe wheat crown rot, indicating that the occurrence of crown rot altered the fungal community structure in the rhizosphere soil. Furthermore, RDA revealed that various environmental variables had distinct effects on the overall fungal community composition. Available phosphorus (r² = 0.936; p = 0.009) and pH (r² = 0.846; p = 0.004) had highly significant impacts on soil fungal community composition (p < 0.01), while wheat crown rot (r² = 0.821; p = 0.017) and available potassium (r² = 0.926; p = 0.027) had significant effects (p < 0.05) (Supplementary Table S2). VPA analysis showed that the occurrence of wheat crown rot and changes in soil chemical properties accounted for 77.51% of the variation in fungal diversity in the rhizosphere soil, with wheat crown rot explaining 35.61%, soil chemical properties explaining 74.01%, and their interaction explaining 32.11%. Among soil chemical properties, available phosphorus had the highest explanatory power at 32.34% (Figure 4D).

Based on the KEGG database, the PICRUSt2 tool was used to functionally predict bacterial 16S OTU information, identifying six primary metabolic pathways: metabolism, genetic information processing, environmental information processing, cellular processes, organismal systems, and human diseases. Within the metabolism pathway, 11 Level 2 functional groupings were identified (Figure 5A). The relative abundances of these 11 metabolic genes showed no significant differences between fields with light (L) and moderate (M) wheat crown rot incidence. However, compared to L, fields with severe (S) wheat crown rot had higher relative abundances of genes associated with xenobiotic degradation and metabolism, amino acids, secondary metabolite synthesis, and carbon metabolism, decreasing by 6.16, 3.24, 3.88, and 3.36%, respectively. In contrast, the relative abundances of genes related to glycine synthesis and signal transduction significantly increased, with reductions of 2.51 and 12.12%, respectively, compared to L.

According to functional predictions by FUNGuild, fungal functional classifications in the samples were identified (Figure 5B), showing that the fungal community could be divided into seven categories: plant pathogens, animal pathogens, dung saprotrophs, undefined saprotrophs, endophytes, mycorrhizal fungi, and unclassified fungi. The relative abundance of plant pathogens showed no significant difference between fields with light (L) and moderate (M) wheat crown rot incidence. However, in field with severe (S) wheat crown rot, the relative abundance of plant pathogens increased by 17.22 and 18.66% compared to L and M, respectively. The relative abundance of dung saprotrophs and undefined saprotrophs gradually decreased with the severity of wheat crown rot. Compared to L and M, the relative abundance of dung saprotrophs in S fields decreased by 25.25 and 7.92%, respectively, while that of undefined saprotrophs decreased by 25.28 and 21.60%, respectively. These results indicate that wheat crown rot increases the abundance of plant pathogens in the wheat rhizosphere while reducing the abundance of saprotrophic fungi.

Of 128 bacterial strains isolated from wheat rhizosphere soil, BF-237 showed the largest zone of growth inhibition (9 mm). BF-237 colonies were milky white, convex, round, with uneven edges, wet surface and large folds on LB agar at 24 h at 37°C. BF-237 clustered with Bacillus velezensis based on genome comparisons with 11 other Bacillus species (Supplementary Figure S1), BF-237 also had the highest ANI value of 98.69% with B. velezensis OR683507.1. Thus, BF-237 was identified as a strain of B. velezensis. The greenhouse experiment showed that BF-237 had 56.61% control effect on wheat crown rot and 14.18% growth promotion effect on wheat in sterilized soil (Figure 6A). While, in natural soil BF-237 had 53.32% control effect on wheat crown rot and 11.50% growth promotion effect on wheat (Figure 6B).