DT (Figure 1A) and HD (Figure 1B) sampling sites were located in Ludian County, Zhaotong City, Yunnan Province, China. ME (Variety name: Liu-mei, obtained from Longxing Biotechnology Co., LTD, China.) was cultivated in October 2021 and harvested in January 2022. The yield of ME in DT and HD filed was 504 kg/km² and 210 kg/km².

We analyzed the pH and abundances of OM, N, P, and K in both DT and HD soils under ME cultivation. In DT soils, no significant differences were observed between the CK and YDJ treatments in the pH (6.4–6.6; Figure 2A), content of OM (45–50.5 g/kg; Figure 2B), and content of N (189.6–201.1 mg/kg; Figure 2C). However, the content of P (15.5–20.1 mg/kg; Figure 2D) and K (170.9–203 mg/kg; Figure 2E) was significantly higher in the YDJ treatment than in the CK treatment (P content: p = 0.0015; K content: p = 0.0193). Therefore, although no significant differences in pH, OM, and N were observed between the CK and YDJ treatments in DT soils, the content of P and K was markedly higher in the YDJ treatment than in the CK treatment, indicating that ME cultivation increased the availability of these nutrients in the soil. No significant differences in the pH of HD soils (5.1–5.2; Figure 2F) were observed between the CK and YDJ treatments. However, the content of OM (55.5–64.9 g/kg; Figure 2G), N (221.8–295 mg/kg; Figure 2C), and P (144.3–206.6 mg/kg; Figure 2I) was significantly higher in the CK treatment than in the YDJ treatment (OM content: p = 0.011; N content: p = 0.0029; P content: p = 0.0017), and the content of K (123–426.2 mg/kg; Figure 2J) was significantly higher in the YDJ treatment than in the CK treatment (p = 0.0046). There were no significant differences in pH in HD soils between the CK and YDJ treatments, and the content of OM, N, and P was markedly higher in the CK treatment than in the YDJ treatment; the content of K was higher in the YDJ treatment than in the CK treatment, suggesting that the effects of ME cultivation on the availability of nutrients soil vary among soil types. Previous studies have indicated that the cultivation of ME with peach trees and ME fermentation with cultivated maize can increase the content of effective K in the soil (Phanpadith et al., 2020; Song et al., 2021), which is consistent with the results of this study. Similar findings were observed in DT and HD soil, and the effect in HD soil was significant. This indicates that ME cultivation can be used as a replacement for some of the positive effects of K fertilizer. Previous studies have shown that interactions between large fungi and soil can increase soil fertility (Duan et al., 2022a; Duan et al., 2023a), which stems from the fact that interactions of fungi with soil can alter chemical properties, soil enzyme activities, ion concentrations, and other factors; stress resistance can be enhanced via interactions with plants (Duan et al., 2022a; Duan et al., 2022b; Duan et al., 2023a). Overall, the results of our study indicate that the cultivation of ME during the fallow periods in DT and HD not only increases the utilization efficiency of farmland but also possibly improving soil nutrient levels.

We conducted a metabarcoding survey to characterize the diversity of soil bacteria under ME cultivation. 16S rRNA sequencing yielded 1,450,023 effective metabarcoding tags. A total of 3,202 bacterial OTUs were identified per sample on average according to OTU clustering analysis of the soil samples (Table 1). A Venn diagram of the OTUs revealed that 726 (10.56%) of the 6,874 bacterial OTUs were present in all soil samples, and 1,012 (14.72%) and 934 (13.58%) of the OTUs were unique to the DT-YDJ and HD-DYJ treatments, respectively (Figure 3A). NMDS plots based on the abundances of OTUs revealed separation in bacterial OTUs between the DT-CK/YDJ (Figure 3B) and HD-CK/YDJ FR (Figure 3C) treatments, indicating that the distribution of bacteria in these treatments significantly differed. We also compared the alpha diversity among samples using a t-test. The mean values of the Sob index of soil samples were 3,723 and 3,419 in the DT-CK and DT-YDJ treatments, respectively (Figure 3D); the mean values of the Sob index were 2,705 and 2,691 in the HD-CK and HD-YDJ treatments, respectively (Figure 3G). The mean values of the Shannon index were 9.8 and 9.58 in the DT-CK and DT-YDJ treatments, respectively (Figure 3E); the mean values of the Shannon index were 8.89 and 9.06 in the HD-CK and HD-YDJ treatments, respectively (Figure 3H). All values of the Good’s coverage index were greater than 0.98, which indicates that the sequencing depth of all samples was sufficient. However, the effect of ME cultivation on the alpha diversity index of soil bacteria was weak in both DT and HD soil, and none of the differences between all comparison groups were significant (p > 0.05). A previous metabarcoding study of soil under ME cultivation at different levels of maturity revealed that around 450 of OTUs were identified, and the Shannon diversity index of soil samples was between 4.8 and 5.1 (Orlofsky et al., 2021). The number of OTUs and Shannon index values were much lower in our study than in this previous study. The Shannon index values of soil at different maturity levels ranged from 4 to 6 in another metagenomic sequencing study of soil under ME cultivation at all maturity stages, and this significantly differs from the results of our study (Zhang et al., 2023). This might stem from differences in soil type among studies; it also suggests that OTU richness and alpha diversity might not be the primary factors underlying the interactions between ME and soil. The fact that we did not detect any differences in alpha diversity between soils under ME cultivation and control soil (i.e., YDJ and CK treatments, respectively) provides additional support for this finding.

We analyzed differences in the composition of the bacterial community between DT and HD soils under ME cultivation based on the SILVA database. The five most abundant bacterial phyla (Figure 4A) were Proteobacteria (28.99%, 24.68%, 26.97%, and 28.01% in the DT-CK, DT-YDJ, HD-CK, and HD-YDJ treatments, respectively), Actinobacteria (8.07%, 9.07%, 12.96%, and 14.57% in the DT-CK, DT-YDJ, HD-CK, and HD-YDJ treatments, respectively), Acidobacteria (9.48%, 12.62%, 9.88%, and 12.15% in the DT-CK, DT-YDJ, HD-CK, and HD-YDJ treatments, respectively), Chloroflexi (12.87%, 15.44%, 6.39%, and 5.41% in the DT-CK, DT-YDJ, HD-CK, and HD-YDJ treatments, respectively), and Planctomycetes (6.82%, 6.65%, 10.12%, and 9.11% in the DT-CK, DT-YDJ, HD-CK, and HD-YDJ treatments, respectively); the five most abundant bacterial genera (Figure 4B) were Sphingomonas (2.15%, 1.66%, 4.15%, and 6.04% in the DT-CK, DT-YDJ, HD-CK, and HD-YDJ treatments, respectively), Gemmatimonas (2.12%, 3.46%, 2.94%, and 3.19% in the DT-CK, DT-YDJ, HD-CK, and HD-YDJ treatments, respectively), Bryobacter (1.98%, 2.85%, 1.94%, and 2.56% in the DT-CK, DT-YDJ, HD-CK, and HD-YDJ treatments, respectively), Flavobacterium (4.8%, 0.95%, 0.08%, and 0.76% in the DT-CK, DT-YDJ, HD-CK, and HD-YDJ treatments, respectively), and Anaeromyxobacter (1.85%, 2.89%, 0.08%, and 0.76% in the DT-CK, DT-YDJ, HD-CK, and HD-YDJ treatments, respectively). LEfSe analysis revealed that Acidobacteria (log₁₀ LDA score >3, p < 0.05) and Gemmatimonadetes (log₁₀ LDA score >3, p < 0.05) were the marker phyla in the DT-YDJ vs. DT-CK comparison group (Figure 4C). Cyanobacteria (log₁₀ LDA score >3, p < 0.05) and Chlamydiae (log₁₀ LDA score >3, p < 0.05) were identified as the marker phyla in the HD-YDJ vs. HD-CK comparison group (Figure 4D). Acidobacteria was enriched in both DT and HD soils under ME cultivation (i.e., DT-YDJ and HD-YDJ treatments); it was also the most abundant bacterial phylum in the DT-YDJ treatment, suggesting that ME cultivation has the most pronounced effect on the abundance of Acidobacteria. Acidobacteria is one of the most common bacteria in soil. It can survive in different soil environments; play a key role in maintaining the structure and function of soil microbial communities, and mediate the conversion of soil nutrients (Lauber et al., 2009; Kielak et al., 2016). Thus, the cultivation of ME in DT and HD soils can enhance soil health by promoting the proliferation of Acidobacteria. Next, we identified bacterial genera that showed significant differences in abundance between DT and HD soils using t-tests. In DT soils (Figure 4E), Bryobacter (relative abundance of 1.98% and 2.85% in the DT-CK and DT-YDJ treatments, respectively; p = 0.04), Anaeromyxobacter (1.85% and 2.89% in the DT-CK and DT-YDJ treatments, respectively; p = 0.01), Candidatus_Solibacter (1.20% and 1.82% in the DT-CK and DT-YDJ treatments, respectively; p = 0.01), Ellin6067 (0.70% and 0.92% in the DT-CK and DT-YDJ treatments, respectively; p = 0.01), and Bacillus (0.15% and 0.56% in the DT-CK and DT-YDJ treatments, respectively; p = 0.001) were the five most differentially abundant genera; in HD soils (Figure 4F), Acidibacter (relative abundance of 0.64% and 0.30% in the HD-CK and HD-YDJ treatments, respectively; p = 0.04), Blastococcus (0.16% and 0.37% in the HD-CK and HD-YDJ treatments, respectively; p = 0.01), and Mesorhizobium (0.10% and 0.27% in the HD-CK and HD-YDJ treatments, respectively; p = 0.04) were the three most differentially abundant genera. In which, a study has shown that Bryobacter and Candidatus_Solibacter in rhizosphere soil has the potential to regulate Cinnamomum migao metabolism (Li et al., 2023); Ellin6067 as a kind of nitrosifying bacteria, which may participate in the soil nitrogen cycling process (Lv et al., 2022); Bacillus is a well-known plant growth-promoting bacterium (Sheng, 2005; Ali et al., 2021). Blastococcus with potential capacity improve the soil nutrient supply (Li et al., 2022); Mesorhizobium be found with capacity of improves chickpea growth under chromium stress and alleviates chromium contamination of soil (Naz et al., 2023). Our findings indicate that the composition of the bacterial community differed significantly in DT and HD soils under ME cultivation. The relative abundances of the top bacterial phyla and genera differed between the two soil types, suggesting that the soil type has a major effect on the bacterial community.

To understand the effects of ME cultivation on the ecological functions of soil, we analyzed the functions of soil bacteria using the OTU data. We used the FAPROTAX and Bugbase databases to predict the ecological functions of microbial OTUs identified in the DT-YDJ and HD-YDJ treatments. We found that the abundances of multiple functional modules related to methane synthesis decreased in DT-YDJ soil, including methanotrophy, acetoclastic_methanogenesis, and methanogenesis_by_disproportionation_of_methyl_groups. Rice cultivation produces a large amount of methane, which results in excessive greenhouse gas emissions (Yan et al., 2005); this result indicates that ME cultivation has the potential to reduce the methane emissions of DT soils. The abundances of functional modules related to nitrification were higher in HD-YDJ soil than in HD-CK soil, including aerobic_ammonia_oxidation, aerobic_nitrite_oxidation, and nitrification. Nitrification is the key aerobic process of the biogeochemical N cycle (Koch et al., 2015). These findings indicate that the cultivation of ME in HD soil can promote soil micronutrient cycling. The BugBase database annotations (Figure 5B) were used to predict the biological functions of microorganisms (Ward et al., 2017). The cultivation of ME in DT soils decreased the abundance of the Facultatively_Anaerobic and Potentially_Pathogenic classes. This indicates that the cultivation of ME in DT soil might lower the severity of rice diseases by reducing the abundance of the Potentially_Pathogenic class, and additional studies are needed to confirm this possibility. The cultivation of ME in HD soils increased the abundance of Stress_Tolerant, which suggests that ME cultivation might related to enhance the resistance of corn growth to stress. Overall, ME cultivation had positive effects on the health of both DT and HD soils.

We mapped the bacterial interaction network according to the Pearson correlation coefficients to clarify the interactions between bacterial genera in DT and HD soils (Figure 6). In DT soils, the network comprised 96 nodes, and three bacterial genera with high abundance and high connectivity (abundance >2% and connectivity >20) were identified, including Gemmatimonas, Bryobacter, and Anaeromyxobacter (Figure 6A). ME cultivation might affect the abundances of bacterial genera in DT soil, which in turn affects the distribution of soil bacterial diversity. Members of the genus Gemmatimonas are capable of anaerobic photosynthesis (Mujakić et al., 2022); however, their effects on soil chemistry remain unclear. The high abundance and connectivity of Gemmatimonas in anaerobic rice soil might indirectly affect soil health by affecting other bacteria. The genus Bryobacter shows chemoorganotrophic activity and can use sugars, polysaccharides, and organic acids as energy sources; it thus plays a key role in soil metabolism (Pertile et al., 2021). Anaeromyxobacter can perform N fixation in soil (Masuda et al., 2020). The network in HD soils consists of 89 nodes (Figure 6B); no bacterial genera with high abundance and high connectivity were detected, and only Chujaibacter was found to have a moderate level of connectivity and abundance (abundance >1% and connectivity >10). ME cultivation in DT and HD soils might affect the abundances of these bacteria and thus soil bacterial diversity and health.

Correlation analyses were conducted to clarify the associations between the distribution of microbes and the content of soil chemicals. First, we analyzed the environmental contributions of soil chemical factors in DT and HD soils to soil microbial diversity following ME cultivation. The contribution rates of P and K in DT soil were high (contribution rate >10%) (Figure 7A); the contribution rates of OM, N, P, and K in HD soil were also high (contribution rate >10%) (Figure 7B). In DT soil, the abundances of Gemmatimonas, Bryobacter, Anaeromyxobacter, Candidatus_Solibacter, Ellin6067, and Flavisolibacter were significantly correlated with the content of P, and the abundances of Anaeromyxobacter and Candidatus_Solibacter were significantly correlated with K (Figure 7C). In HD soil, the abundance of Acidibacter was significantly correlated with P, and the abundances of Bryobacter, Bacillus, Streptomyces, and Paenarthrobacter were significantly correlated with K (Figure 7D). The abundances of the aforementioned bacterial genera are likely affected by the transformation and accumulation of soil chemical properties. Some of these bacterial genera repeatedly appear in the bacterial interaction network (Figure 6), such as Gemmatimonas, Bryobacter, and Anaeromyxobacter, which emphasizes their importance in regulating soil chemistry processes following ME cultivation. We also identified some bacterial genera that were significantly correlated with the accumulation of K following ME cultivation in HD soil. No studies currently indicate that there is a correlation between the abundance of Bryobacter and the accumulation of K in soil. Thus, further exploration of the mechanism underlying the effects of this bacterial genus on the accumulation of K in soil is needed to maximize the utilization of this bacterial resource and enhance soil health. Bacillus can stimulate the dissolution of soil K and mediate the absorption of K by plants (Sheng, 2005; Ali et al., 2021); ME cultivation in cornfields might increase the content of soil K through the enrichment of Bacillus. Streptomyces is the most abundant and arguably the most important actinomycete genus; it is also a rich source of bioactive compounds, antibiotics, and extracellular enzymes (Olanrewaju and Babalola, 2019). Streptomyces might regulate the content of soil K through its interactions with soil. Members of the genus Paenarthrobacter can play important roles in soil metabolism, as they can degrade various soil pollutants such as herbicides (Zhao et al., 2022). Its metabolic processes might also be related to the accumulation of K, but this possibility needs to be experimentally verified. In general, ME cultivation might affect the abundances of these bacterial genera, and these bacteria might play important roles in regulating soil chemical properties. Additional studies of the interaction between these bacteria and soil are needed. There is a special need for analyses of strains that play a major role in soil K accumulation in Bryobacter, Bacillus, Streptomyces, and Paenarthrobacter; isolation of these taxa could aid the development of new biofertilizers.