3.1 Effects of different treatments on soybean agronomic traits and soybean yield
Four treatments for plant height, root length, and root crude were measured throughout six periods of soybean growth to investigate the effects of compound inoculants and straw return on the agronomic parameters of soybean. Effects on plant fresh weight and biomass. Conventional fertilization had a minimal effect on agronomic traits, with almost no change in the first three periods. In the latter three periods, it improved plant height, stem thickness, and fresh weight, indicating that chemical fertilizer applied alone could still positively affect soybean growth (Table S1). The plots with straw return had the most significant changes in long-root coarseness for soybean plant height, indicating that straw plays an important role in promoting overall soybean growth. The addition of composite fungicides was particularly noticeable in soybean above- and below-ground biomass and plant fresh weight alterations. The combined effect of straw-return fields and the composite fungicide on plant fresh weight and external shape was most significant, indicating that the combined effect was more significant than the single effect. Among the four investigated treatments, the soybean drum stage differed the most. Compared to F, each soybean treatment boosted wheat yield from 3.91% to 15.46%, with the CSF treatment showing the most significant increase. The yields of the varied treatments were expressed as F > CF > SF > CSF, with significant differences.
3.2 Effects of different treatments on the physicochemical properties and nutrient content of soybean
Table 1 shows the basic physicochemical properties of the soil that were significantly affected by the different fertilization treatments. The table shows that the content of each treatment varied throughout the reproductive period. The soil pH increased continuously with the addition of composite mycelium and straw, and the SOM steadily reduced, with the composite mycelium reducing effect is higher than the straw return effect. The water content gradually increased, and the organic C showed an initial rising. Then declining trend, lower than that of F. AK, gradually decreased. Finally, AP gradually increased, with the CSF effect being more significant than that of SF and CF, with significant variations across the six treatments.
Table 1
3.3 Effect of different treatments on soil enzyme activities
Fig. 1 depicts the activities of four enzymes (sucrase, catalase, urease, and acid phosphatase) at various soybean growth stages. The rhizosphere soil enzyme activities increased as the soybeans grew and developed under varied treatments. Overall, the activities of the four enzymes were higher with composite fertilization than that with individual fertilization. The enzyme activities in the CSF treatment showed the highest trend, with an increase in SF > CSF > CF > F, and there were significant differences between treatment groups (P < 0.05). In soybean, catalase activities increased from 10.19%–15.71% in the mature stage compared to the seedling stage, with SF > CSF showing insignificant differences at the seedling stage. In sucrase, the mature stage increased from 25%–30.30% compared to the seedling stage. Among the ureases, a 23.53%–40.62% increase was observed at maturity compared to the seedling stage, with SF > CSF showing significant differences at the soybean outplanting stage. The mature stage of acid phosphatases showed a 5.68%–14.86% increase compared to the seedling stage, with significant differences across the emergence and branching stages SF > CSF > CF.
Fig. 1
3.4 Number of soil microorganisms cultivated under laboratory conditions from soybeans under different treatments
Supplementary Fig. 1 depicts the numbers of soil microorganisms cultured under laboratory conditions for soybeans under various treatments. Under all four treatments, soil bacterial populations increased initially before decreasing. The bacterial numbers in each treatment peaked at the pod stage, where the f treatment had the lowest numbers, the CSF treatment had the highest numbers, SF was approximately the same as the CF treatment, and the mature soybean stage had a lower number than the emergence stage, with a 3.12%–5.15% decrease. Except for the CF treatment, all the other three treatments showed an initial increase and a decreasing trend in soil fungal populations. The CF treatment showed a decreasing trend: CSF > CF > SF > F, the number of soybean maturity stages was lower than the emergence stage, and the decreasing amplitude was 6.98%–11.37%. The various treatments roughly showed a decreasing trend first, followed by an increase and then a decrease, and the number of soybean mature stages was lower than that at the emergence stage, with a decrease of 21.28%–46.67%, with the four treatments at anthesis having the highest number. The CSF treatment reached 180 × 105 / g.
3.5 Alpha-diversity of microbial community
For six months, 24 samples were collected from the four soybean treatments, and the soil bacterial coverage (coverage) index ranged from 0.9098–1.000. The amount of sequencing data was reasonable, and the sequencing data covered all bacterial taxa in the soil, which can significantly reflect the bacterial properties in the soil environment. Using Illumina MiSeq high-throughput sequencing, all samples yielded 2,075,098 optimized sequences, with 500,737–533,418 optimized sequences obtained per sample. The obtained data were classified at a 97% similarity level, which detected 29,729 out of 1,834, 1,380, 2,019, and 2,132 unique to F, CF, SF, and CSF, respectively. Among them, the concentration of bacteria in cerebrospinal fluid (CSF) increased by 16.25% compared to that observed in the F treatment. In comparison, the bacterial count in CF decreased by 32.90% compared to the F treatment. Qime was used to calculate Chao1, an indicator of flora richness, and Simpson, an indicator of flora diversity. At the phylum level, the lowest Chao1 index was found in SF, which decreased by 45.45% compared to that in F, and the Chao1 index in CSF, which decreased by 15.00% compared to that in CF, all without statistical significance. The lowest SimSpon index was found in CF, with a 2.1% decrease compared to F, and the highest in SF, with no statistical significance. At the genus level, the lowest Chao1 index was found in SF, which decreased by 14.23% compared to that in F, and the Chao1 index of CSF decreased by 3.38% compared to that in CF, with no significant differences. The CSF had the lowest Simplon index, with a contrast reduction of 9.25%, neither of which was significant. In conclusion, incorporating straw-returned fields into fertilized croplands reduced Soil Rhizosphere flora richness while increasing community diversity. At the genus level, adding a compound fungicide to fertilized croplands reduced flora diversity while increasing flora richness. Complex fungicides in straw-returned fields significantly reduced community diversity at the phylum level. These findings suggest that compound fungicides enhance straw return while inhibiting soil microbial growth.
Table 2
3.6 Effects of different treatments on soil bacterial community composition and soybean abundance
A bacterial community phylogenetic analysis of genes was performed to identify 20 phyla with the highest relative abundances. At the phylum level, 16 bacterial phyla were detected, with acidobacter, Proteobacteria, Chloroflexi, Bacteroidetes, and Actinobacteria being the dominant phyla, accounting for 85.79% of all sequences, and relative abundances of 10.79%–41.12%, 21.30%–36.70%, 2.21%–13.91%, 3.46%–8.29%, and 2.43%–8.90%, respectively (Table S1). Some less abundant phyla, such as gemmatimonadetes, Firmicutes, verrucomicrobia, wps-2, armatimonadetes, patescribacteria, and planctomycetes, nitrospirae, were detected in the various treatment samples during the six soybean periods, with relative abundances ranging from 0.014%–4.8%, with an additional 0.88% being unclassified bacteria (Fig.2a). The relative abundances of acidobacter and Chloroflexi were higher in the CF treatment than in the other three treatments, with Proteobacteria in the SF, Bacteroidetes in the CSF, and Actinobacteria in the F treatment having the highest relative abundances. At the genus level, 18 bacterial genera were detected, with the dominant genera being Candidatus_ Solibacter, Sphingomonas, bryobacter, gemmatimonadota, and Burkholderia caballeronia paraburkholderia, accounting for 35.88% of all sequences, and with relative abundances of 7.04%–13.10%, 2.24%–13.9%, 1.95%–15.79%, 0.89%–3.68%, and 015%–8.14%, respectively. Certain less abundant genera, such as mucilaginibacter bacillus, hsb-of53_ F07, and granulicella, were also detected in the various treatment abundances ranging from 0.015%–2.37%. The genus Bacteroidetes contained a high proportion of unclassified bacteria (15.02%–50.80%), resulting in a less dominant flora (Fig.2b). Bryobacter relative abundance was higher in CF than in the other three treatments. The relative abundance of Sphingomonas was the highest in SF, while the relative abundance of Candidatus _ Solibacter was the highest in CSF.
Fig.2
3.7 Beta-diversity of microbial community
Non-metric multidimensional scaling (NMDS) analysis of species based on the Bray–Curtis distance algorithm (fig.2c-d) revealed that the distribution of bacterial communities differed across the treatments. Visual analysis using NMDS lead to the classification of the bacterial communities into four distinct groups at the phylum level (fig.2c), with the stress of 0.077, < 0.1 performing better. There were noticeable variations in the community structure between each epoch of each sample, and the nmds1 axis shows a clear distinction between the communities of groups F and CSF, with the CSF > F and SF span containing the F and CF groups. The nmds2 axis shows a significant distinction between the three groups: CF, SF, and CSF, with CF > CSF = f > SF. This finding indicated that under fertilization settings, community importance was significantly elevated in the case of the combined fungicide and straw-returned fields, with the combined fungicide alone being higher than in the straw-returned fields. Thus, the bacterial community structure became unstable and altered due to CSF application. At the genus level (Fig.2d), the bacterial communities were classified into three distinct groups, with stresses of 0.155 and < 0.2, indicating the reliability of the data. There was a noticeable differential in the community structure between each sample epoch, as seen from the nmds1 axis. For CF vs. SF, the community of the CSF group was clearly distinguished, where CF > SF = CSF. The CF connected the SF with the CSF along the NMDS2 axis, where CSF > SF. These findings indicate that, under fertilization conditions, the addition of the composite fungicide modified the soil bacterial community structure, and the composite fungicide mixed with straw functioned better than straw alone.
3.8 Differential analysis of microbial communities of soybeans under different treatments
Linear Discriminant Analysis (LDA) was performed for the four treatments, revealing that Firmicutes were enriched in treatment F at the phylum level. At the genus level, bacterium_ Eillin5129 was abundant in the SF treatment, Bacillales in the F treatment, devosia and caulobacteraceae in the CSF treatment, and camdidatus_ Koribacter in the CF treatment (Fig.3a-b)
Fig.3
3.9 Functional prediction analysis in soybean
Kyoto Encyclopedia of Genes and Genomes (KEGG) was used for detection. Sugar metabolism (carbohydrate metabolism), polyketones (metabolism of terpenoids and polyketides), amino acid metabolism, calcium factor and station compounds, vitamin metabolism (metabolism of cofactors and vitamins), energy metabolism, other acid metabolism (metabolism of other amino acids), lipid metabolism, replication and repair, biodegradation and metabolism (xenobiotic biodegradation and metabolism), and polysaccharide synthesis and metabolism are all part of the second pathway. The content differed slightly between the treatments. Attachments can be seen in the third pathway affecting sugar metabolism in the second pathway, with the most prevalent being ansamycins, biosynthesis of vancomycin antibiotics, anabolic leucine, and isoleucine biosynthesis.
3.10 Correlation analysis between soil physicochemical properties and soil microbial communities
Changes in bacterial phyla were negatively connected with AN, positively correlated with pH, and negatively correlated with AK (P < 0.01), according to a Mantel test examination of the association between bacterial and fungal populations. Changes in bacterial genera correlated positively with SOC, while SOM correlated positively with AN and negatively correlated with AP (Table S2).
Redundancy analysis (RDA) was used to analyze the relationships between various treatments and soil physicochemical factors (pH, SOC, SOM, AN, AK, and AP). Fig. 4a–b shows the results. Among the phylum-level treatments, different treatments exhibited significant positive or negative correlations with the selected environmental factors: pH (0.00128 of explained total variance), SOC (0.00432 of explained total variance), SOM (0.04289 of explained total variance), AN (0.00294 of explained total variance), AP (0.103 of explained total variance), and AK (0.4588 of explained total variance). The selected factors accounted for 91.6% of the total variations in bacterial community changes. At the genus level, there were significant positive and negative correlations between pH (0.00384 of the total variance explained), SOC (0.00793 of the total variance explained), SOM (0.03976 of the total variance explained), AN (0.02454 of the total variance explained), AP (0.05484 of the total variance explained), and AK (0.15638 of the total variance explained). The factors selected accounted for 73.5% of the total variations in bacterial community changes.
Fig.4
3.11 Pearson analysis of soil physicochemical properties and microbial communities
To explore the effects of soil physicochemical properties on bacterial communities, the link between the physicochemical parameters of black soil and the levels of phyla and genera was examined in this study (Fig. 4c-d). The results showed that soil AP at the phylum level was significantly correlated with Dependentiae; WPS-2, tenericutes, epsilonbacteraeota, elusimicrobia, and omnitrophicaeota were significantly correlated with latescibacteria; AK was significantly correlated with Proteobacteria, and elusimicrobia was significantly correlated with pH; fibrobacteria, dependentiae, cyanobacteria, SOC, and SOM were significantly correlated with elusimicrobia and omnitrophicaeota. At the genus level, soil SOM and Candidatus_ Koribacter were significantly correlated; AP was significantly correlated with massilia, jatrophihabitans, Bradyrhizobium, halingum, niastella, ellin6067, acidibacter, Pseudomonas, and Stenotrophomonas; AK was significantly correlated with actinospica, and Candidatus_ Koribacter, ellin6067 were significantly correlated; and pH was significantly correlated with catenulispora, acidibacter, niastella, phylobacterium, and jg30a-kf-32.
3.12 Pearson correlation analysis between black soil physicochemical properties and soybean yield
Table 3 shows the correlation analysis between black soil physicochemical properties and soybean yield under the four treatments. There was a significant positive correlation between soil pH and F, a significant negative correlation between soil SOM and SF, and soil moisture content CF and SOM. Further, there was a significant positive correlation between SOC and CSF and between AK and F, a significant negative correlation with SF, positive correlation between AP and CF. Finally, there was a highly significant negative correlation with CSF but no significant correlation between AN and soybean yield. Pearson's correlation analysis revealed that MC, AP, and AK were significantly correlated with soybean yield, indicating that an appropriate increase in phosphorus and K fertilizers were beneficial to soybean yield and may be the most important environmental factor affecting soybean yield.
Table.3