Rhizosphere soil nutrients
Table 1 illustrates the impact of disparate fertilisation techniques on rhizosphere soil nutrients. Significant differences were observed in the effects of different fertilisation measures on rhizosphere soil nutrients (P < 0.05). Compared with the control (CK), all fertilization treatments significantly increased rhizosphere soil pH and nutrient levels, including organic matter (OM), Total potassium (TK), Total phosphorus (TP), Total nitrogen (TN), Available potassium (AK), Available phosphorus (AP), and Hydrolyzable nitrogen (HN). As the proportion of organic fertilizers in use has increased, there has been an initial rise in rhizosphere soil pH, OM, TN, TP, and HN, which has subsequently declined. The highest values were observed in the CD3 treatment, with a pH of 6.26, OM content of 36.23 g/kg, TN of 1.58 g/kg, TP of 0.81 g/kg, and AN of 160 mg/kg. Compared to NPK trial, soil pH, OM, TP, AP and HN in CD3 exhibited a marked increase by 4.33%, 18.67%, 20.90%, 23.35% and 32.97%, respectively. The results demonstrate that the application of high levels (75%) of organic fertilizer in place of chemical fertilizer can enhance the rhizosphere soil nutrient content in Albic soil crops and mitigate soil acid-base imbalances.
Rhizosphere soil microbial diversitys
Diversity data were analyzed for 18 soil samples. The 16 S rRNA and ITS gene sequencing results yielded a total of 1,126,615 and 1,191,087 valid sequences, respectively. After OTU clustering at 97% similarity, bacteria yielded a total of 1,087,621 optimized sequences, comprising 452,900,350 bases with an average sequence length of 416 bp; fungi yielded a total of 1,132,889 optimized sequences, comprising 282,519,897 bases with an average sequence length of 249 bp. The sequencing saturation and depth for all samples satisfied the experimental requirements.
The Shannon, Ace, and Chao1 indices were employed to evaluate the impact of disparate fertilisation treatments on microbial diversity in rhizosphere soil. The findings demonstrated that the implementation of fertilisation techniques resulted in a notable impact on the diversity of bacterial and fungi communities (P < 0.05). High levels of organic fertilizer replacing chemical fertilizer treatments significantly increased bacterial diversity and richness but decreased fungi diversity and richness. In particular, the bacterial ACE and Chao1 indices of CD1, CD2, and CD3 treatments were greater than those in NPK, And CD3 treatment had the significant increase of ACE and Chao indices by 19.49% and 21.02%, respectively (Fig. 1 C and E). Meanwhile, the Shannon index was significantly increased by 3.38% in CD3 only. (Fig. 1 A). For fungi, however, the Shannon, ACE, and Chao1 indices of the CD3 treatment were significantly lower than those of the NPK treatment by 15.53%, 34.05%, and 32.43%, respectively (Fig. 1B、D and F).
Shannon index (A, B), ACE index (C, D) and Chao 1 index (E, F), bacteria and fungi, and PCoA analysis based on Anosim intergroup difference test (G. bacteria and H. fungi) for different fertilization treatments. Data are means ± SE (n = 3). Different letters indicate significant differences (P < 0.05) according to ANOVA.CK: no fertilizer application (CK); NPK: Conventional fertilization; CD1, CD2 and CD3 are 25%, 50 and 75% treatments of organic fertilizers replacing chemical fertilizers; CD4 is 100% organic fertilizer replacing chemical fertilizers.
The results of PCoA demonstrated that there were significant differences in the bacterial and fungi communities between the six treatments (P = 0.001). The rhizosphere soil bacterial and fungi communities of the NPK, CD1, and CD2 treatments exhibited greater similarity, while those of the CD3, CD4, and CK treatments were more distinct from the others. The bacterial and fungi communities in CD3 treatment showing a clear separation trend from those in NPK treatment. It indicates that fertilization measures can change the rhizosphere soil bacterial and fungi community structure. And replacing chemical fertilizer with high ratio of organic fertilizer has a particularly significant effect (Fig. 1G and H).
Rhizosphere soil microbial community composition and structure
Cluster analysis was performed using the USEARCH-uparse algorithm with 97% OTU sequence similarity and 0.7 classification confidence. The SILVA 138/16S_Bacteria database was used for bacterial classification, and the UNITE 8.0/ITS_Fungi database was used for fungi species classification, with sequences from one or more samples grouped together based on these criteria. As shown in figure. S1, the six treatments shared 1788 OTUs for rhizosphere soil bacteria and 227 OTUs for fungi. Bacterial OTU numbers were in the order of CD3 > CD4 > CD2 > CD1 > NPK > CK. And fungi OTU numbers followed the order of CD2 > NPK > CD3 > CD4 > CD1 > CK. The annotation results showed that bacterial OTUs belonged to 41 phyla, 142 orders, 566 families, and 1112 genera, while fungi OTUs belonged to 14 phyla, 53 orders, 116 families, and 496 genera.
To investigate the impact of replacing chemical fertilizer with organic fertilizer on rhizosphere soil bacterial and fungi communities in soybean grown in Albic soil, comparative analyses were conducted at the phylum and genus levels. The results indicated that treatments involving organic fertilizer (CD1, CD2, CD3) influenced the abundance of soil bacteria and fungi at various taxonomic levels. At the phylum level, the top 10 dominant bacterial and fungi species in rhizosphere soil showed consistent abundance, although there were differences among specific species. Among the bacterial phyla with high relative abundance (> 5%) in all treatments were Proteobacteria (21.07–30.11%), Actinobacteriota (15.94–27.74%), Acidobacteriota (9.61–14.59%), Chloroflexi (9.58–14.59%), and Verrucomicrobiota (5.03–6.95%) (Fig. 2 A). The relative abundance of Proteobacteria exhibited a marked increase in the CD1, CD2, CD3, and CD4 treatments than in NPK, by 35.45%, 20.45%, 36.11% and 42.90%, respectively. While a decrease (as evidenced in Table S1) in the relative abundance of Acidobacteriota, Chloroflexi, and Verrucomicrobiota was observed. For fungi, Ascomycota, Basidiomycota, Mortierellomycota, k__Fungi, and Chytridiomycota were the phyla with higher relative abundance (> 5%). Compared with the control (CK), all fertilizer treatments reduced the relative abundance of Ascomycota, while the opposite was true for Basidiomycota, Mortierellomycota, and k__Fungi. The CD3 treatment had the highest relative abundance of Basidiomycota and the lowest of Ascomycota, increasing by 286.79% and decreasing by 26.28% compared to NPK, respectively (Fig. 2B and Table S2).
Bacterial and fungi community composition and structure of rhizosphere soil in different treatments. (A), (B), (C) and (D) represent the phylum and genus levels of bacteria and the phylum and genus levels of fungi, respectively. The bar chart shows the top ten species in relative abundance at the phylum level, with the rest represented by Others, and the heat map shows the top 30 species in relative abundance at the genus level, with red representing high abundance and blue representing low abundance, and the shade of the color denoting the magnitude of relative abundance.
At the genus level, various fertilization treatments influenced the relative abundance of both bacterial and fungi species in the rhizosphere soil, with fungi species exhibiting more pronounced changes. The predominant bacterial genera across treatments were g_norank_f_norank_o_Gaiellales (3.35–7.75%), g_Candidatus_Udaeobacter (4.12–5.89%), g_Sphingomonas (3.01–6.46%), and g_norank_f_norank_o_Acidobacteriales (2.49–3.94%). Notably, all fertilization treatments led to a decrease in the relative abundance of Sphingomonas, while an increase in the relative abundance of g_norank_f_norank_o_Acidobacteriales (Fig. 2 C and Table S3). Among fungi, the genera Conocybe, Mortierella, Fusarium, k__Fungi, Phoma, Schizothecium, p__Chytridiomycota, and Didymella showed higher relative abundance. Further analysis revealed that the application of organic fertilizer (CD1-CD4) specifically reduced the relative abundance of Fusarium, Phoma, and Didymella, but notably increased that of Conocybe, with the most significant increase observed in the CD3 treatment (Fig. 2D and Table S4).
Taxa enrichment differences between treatment groups were analyzed using LEfSe, with an linear discriminant analysis(LDA) threshold of > 4.0. The results indicated significant differences in enriched markers between bacteria and fungi across treatment groups, and fungi exhibited a greater number of distinct markers. The rhizosphere soil bacterial community exhibited 10 taxa with significant enrichment differences between treatments, with CK, NPK, CD1, and CD3 showing 1, 2, 3, and 4 distinct markers, respectively. Notably, the order Burkholderiales was significantly enriched exclusively in the CK treatment, with an LDA score higher than other taxonomic units (Figure. S2). The orders Sphingomonadales, Sphingomonadaceae, Sphingomonas, Anaerolineae, and SBR1031 were enriched in the CD1 and CD3 treatments, respectively (Fig. 3 A). In the fungi community, a total of 53 taxa showed significant enrichment differences, with 7 major taxa in the CK treatment predominantly belonging to the phylum Ascomycota, order Pleosporales, and genus Didymella, where the phylum Ascomycota also had the highest LDA score (Figure. S3). In the NPK treatment, the number of significantly different markers was the lowest, while CD2 and CD3 treatments had higher counts, with 15 and 11 markers, respectively. The primary enriched taxa included the order Sordariales, phylum Mortierellomycota, family Mortierellaceae, genus Mortierella, order Mortierellomycetes, and phylum Basidiomycota, order Agaricomycetes, order Agaricales, and genus Conocybe. CD4 exhibited a total of seven differentially abundant taxa, primarily within the phylum Glomeromycota and the order Tremellodendropsidales (Fig. 3B). The results indicate that the partial replacement of chemical fertilizers with organic alternatives can influence the composition of bacterial and fungi communities by modulating the abundance of specific bacterial groups.
LEfSe analysis of Rhizosphere soil bacteria (A) and fungi (B) from different fertilization treatments. Different colored areas represent different components (red for CK, blue for NPK, green for CD1, pink for CD2, purple for CD3, and orange for CD4), and the diameter of each circle is proportional to the relative abundance of the taxonomic unit, with inner to outer circles corresponding to phylum to genus levels.
Correlation between microbial community structure and environmental factors
The significant relationship between the major bacterial and fungi communities in the rhizosphere soil and soil nutrients was studied using RDA analysis. Bacterial RDA analysis revealed that the two major axes explained 34.45% and 16.20% of the variance, respectively. Specifically, TK, AK, TP, AN, and TN were positively correlated with Acidobacteriota, Myxococcota, and Firmicutes, predominantly found in the first quadrant Whereas AP, OM, and pH correlated positively with Proteobacteria, Gemmatimonadota, and Bacteroidota, mainly found in quadrant IV. Actinobacteriota showed a negative correlation with all rhizosphere soil nutrients. pH (r²= 0.4843, P = 0.005) and OM (r²= 0.4795, P = 0.009) were the primary environmental factors influencing bacterial communities (Fig. 4 A and Table S5). Fungi RDA analysis indicated that the two major axes accounted for 55.93% and 8.82% of the variance, respectively. Specifically, AN, pH, and TP were positively correlated with Basidiomycota, primarily located in the second quadrant. OM, AP, TN, TK, and AK correlated positively with Glomeromycota, Mortierellomycetes, and unclassified fungi, mainly found in the third and fourth quadrants. Conversely, Ascomycota exhibited negative correlations with all nutrients. OM (r² = 0.5708, P = 0.001) was the key environmental factor influencing fungi communities (Table S6). The findings demonstrate a notable correlation between rhizosphere soil nutrients including pH, OM, and TP, and the structure of bacterial and fungi communities (Fig. 4B). Furthermore, these findings suggest that the composition and structure of these communities may be indirectly influenced by the partial substitution of chemical fertilizers with organic ones, which could enhance rhizosphere soil nutrient levels.
RDA analysis of rhizosphere soil nutrients with bacterial community (A) and fungi community (B). Note: pH: acidity and alkalinity; OM: organic matter; TN: total nitrogen; TP: total phosphorus; TK: total potassium; AP: available Phosphorus; AK: available potassium; HN: Hydrolyzable nitrogen.
Characterization of soil microbial networks under different fertilization treatments
Microbial co-occurrence networks were established to analyze the relationships between soil bacterial and fungi species under various fertilization conditions. The networks for the NPK, CD1, CD2, CD3 and CD4 treatments featured 6049, 6418, 6868, 6842, 6800 edges for bacteria, and 5614, 5397, 5297, 5470 and 4983 edges for fungi, respectively. It indicates that partial replacement of chemical fertilizers with organic ones strengthens bacterial interspecies interactions. The average degree of connectivity networks of NPK, CD1, CD2, CD3 and CD4 trails was 60.794, 64.503, 69.025, 68.420, 68.342, and 56.140, 53.970, 53.236, 54.700 for bacterial and 49.830 for fungi networks, respectively. Network densities of NPK, CD1, CD2, CD3 and CD4 trails were 0.307, 0.326, 0.349, 0.344 0.345 for bacteria, and 0.282, 0.271, 0.269, 0.275 and 0.250 for fungi, respectively (Fig. 5 A and B). Treatments with organic fertilizers consistently demonstrated higher average degrees and network densities than NPK, suggesting an increase in the intensity of bacterial interactions and a denser network structure. Conversely, the impact on fungi networks followed an opposite trend. The relative abundances of bacterial and fungi species within the co-occurrence networks varied across treatments. Proteobacteria, Acidobacteriota and Actinobacteriota exhibited a higher degree of prevalence in bacterial networks. Ascomycota was predominant in the fungi network. Although the organic fertilizer treatment simplifyied the network structure, it did not significantly alter the relative abundance of Ascomycota. The bacterial taxa composition remained consistent across treatments, yet the CD4 treatment displayed the lowest diversity of fungi taxa.
Co-occurrence network of soil OTUs in different fertilization treatments. (A) bacteria, (B) fungi. Each node (circle) represents an OTU. node color indicates assigned gate. The size of the nodes is proportional to the relative abundance of the species.
