Impact of different fertilization treatments on production of Capsicum annuum L.
In comparison to the blank control treatment (CK), the application of organic fertilizer (T1) significantly improved the production of Capsicum annuum L. (P < 0.01). The production yield of T3, which involved the application of organic fertilizer and bacillus subtilis, increased by 170.3% compared to T6 (bacillus subtilis only). Similarly, the production yield of T3 (organic fertilizer and compound fertilizer) improved by 180% compared to T5 (compound fertilizer only). Notably, the highest yield of Capsicum annuum L. was achieved in the T4 treatment, which involved the application of organic fertilizer, compound fertilizer, and Bacillus subtilis. These findings clearly demonstrate that organic fertilizer can effectively promote the yield of Capsicum annuum L.
Soil physical and chemical properties and enzyme activity
The effects of different rates of organic fertilizer, Bacillus subtilis, and compound fertilizer on soil chemical properties were examined. Figure 2 illustrates the results. Without the single application of Bacillus subtilis, all treatments led to an increase in pH in both the rhizosphere soil and topsoil compared to the control treatment (ZQ), although the differences were not significant (p > 0.5, Fig. 2A).
Soil electrical conductivity (EC) is an indicator of water-soluble salts in the soil and can influence crop growth. Different plants have different optimal ranges for soil conductivity. Compared to the control treatment, both the rhizosphere and topsoil EC values were significantly reduced in all treatments (p < 0.05). The T4 treatment exhibited the lowest EC values, with reductions of 34.78% and 45.10% in the rhizosphere and topsoil, respectively (Fig. 2B). It is worth noting that the EC value of the rhizosphere soil was significantly reduced in all treatments except T6. Bacillus subtilis alone had a minimal effect on conductivity.
When compared to the control treatment, the contents of total nitrogen (Fig. 2D), total phosphorus (Fig. 2E), and available potassium (Fig. 2F),were relatively higher in T1-T5. Among the rhizosphere soil samples, the total N content in T4 was lower compared to T1, T2, and T3, while no significant differences were observed among the topsoil samples. Additionally, the available potassium content in the topsoil samples of each treatment was higher than that in the rhizosphere soil samples. Compared with ZQ, cation exchange capacity and content of organic matter of T4 treatment were decreased slightly while no significant differences was observed (p > 0.05, Fig. 2J and 2K). Otherwise, increased cation exchange capacity and content of organic matter of T4 were showed compared with CK.
The soil catalase activity in the rhizosphere soil of the T4 treatment was significantly higher than that in the control treatment (p < 0.01, Fig. 2C), with an increase of 200.0%. There was no significant difference in catalase activity among the topsoil samples of T1 to T5. The T4 treatment exhibited the highest sucrose enzyme activity in both the rhizosphere and topsoil, with increases of 90.9% and 24.2%, respectively, compared to the control soil samples (Fig. 2H). Phosphatase activity in the rhizosphere soil of T1-T6 samples increased compared to the control samples (p < 0.05), while no significant difference was observed in the topsoil (Fig. 2G). Urease activity in the T4 samples was the highest in both the rhizosphere and topsoil, with increases of 36.5% and 43.3%, respectively, compared to the control (Fig. 2I).
Effects of different fertilization treatments on OTUs of bacterial communities
To analyze the bacterial communities, the QIIME software (version 1.8.0) was used with the UCLUST algorithm to perform clustering analysis on the tags at a 97% similarity threshold, resulting in the identification of operational taxonomic units (OTUs). Alpha diversity indices were calculated based on the OTUs. Bacterial diversity was assessed using OTUs. Rarefaction analysis indicated that the OTUs were well represented in all generated libraries. Both the Shannon-Wiener analysis and Good's coverage analysis (with Good's coverage values between 98.91% and 99.57%) confirmed that the sequencing depth was sufficient to capture the true microbial composition in the samples, indicating the success of the deep sequencing identification (Fig. S1 and Fig. S2)
In the rhizosphere soil, a total of 1122 OTUs were identified (Fig. 3A). The different treatments exhibited 1, 2, 7, 2, 1, 0, 9, and 3 unique OTUs in T1, T2, T3, T4, T5, T6, ZQ, and CK, respectively.
The OTU distribution Venn analysis showed that there were a total of 1040 OTUs in the topsoil (Fig. 3B). Among the different treatments, T1, T2, T3, T4, T5, T6, ZQ, and CK had 1, 2, 2, 0, 5, 4, 6, and 0 unique OTUs, respectively.
A: rhizosphere soil samples of each treatment; B:Top soil samples of each treatment
Overall structural changes in rhizosphere and top soil bacterial communities
The relative abundances of the top 10 most abundant bacterial species in each treatment group were analyzed and compared. The bacterial community structures in each fertilizer treatment differed from those observed in the long-term replant soil (Fig. 4). Furthermore, the bacterial community structures in the rhizosphere soil were distinct from those in the topsoil.
Specific changes in relative abundance were observed in certain bacterial species. The relative abundances of Chujaibacter increased in T2, T4, and T6 treatments, while bacterium_o_KF-JG30-C25 showed an increased relative abundance in T1-T6 treatments. Additionally, bacterium_f_JG30-KF-AS9 exhibited an increased relative abundance in T2 and T4 treatments (Fig. 4A). In the topsoil samples, an increased relative abundance of f_Acidobacteriaceae_Subgroup_1 was observed in T3 and T4 (Fig. 4B).
LEfSe analysis of intergroup samples
LEfSe (Linear Discriminant Analysis Effect Size) was employed to identify biomarkers with statistical differences among different groups. The histogram of LDA (Linear Discriminant Analysis) value distribution and the branching diagram of LEfSe analysis are presented in Fig. 5. The graph displays species with LDA scores greater than the set threshold (default value is 4.0). The length of the bar graph represents the magnitude of the impact of different species, while different colors represent species in different groups.
In the rhizosphere soil samples, the main bacteria identified in the CK group were Staphylococcaceae. In T3, the main bacterium was Micropepsaceae, while in T5 it was f_Solibacteraceae_Subgroup_3, o_Solibacterales, g_Candidatus_Solibacter, c_Deltaproteobacteria, and o_Myxococcales. T6 exhibited the main bacterium as c_Gammaproteobacteria, o_Ktedonobacterales, and bacterium_f_JG30_KF_AS9.
In the topsoil samples, the main bacteria identified in the ZQ group were f_Xanthobacteraceae, g_Sphingomonas, p_Bacteroidetes, and c_Bacteroidia. In the CK group, the main bacteria were o_Myxococcales, o_Subgroup_2, s_bacterium_o_Subgroup_2, and f_bacterium_o_Subgroup_2. T1 samples exhibited bacterium_c_Actinobacteria and g_Mizugakiibacter as the main bacteria. In T2, the main bacteria were p_Proteobacteria, g_Mizugakiibacter, and c_Actinobacteria. T4 displayed o_KF_JG30_C25 as the main bacterium. T6 showed c_Gammaproteobacteria, o_Xanthomonadales, and f_Rhodanobacteraceae as the main bacteria.
Network analysis of bacterial communities
To reveal the interaction within the topsoil bacterial communities across the Capsicum annuum L replanting processes, we constructed a co-occurrence network at the genus level in the replant system (Fig. 6). The genus Acidobacteriaceae_Subgroup_1 co-occurred with Micropepsaceae, JG30-KF-AS9, Sphingomonas, KF_JG30_C25, and Chujaibacter. The genus Acidobacteriales co-occurred with Candidatus_Solibacter; Chujaibacter co-occurred with Isosphaeraceae, Reyranella, and JG36_TzT_191. Gammaproteobacteria_Incertae_Sedis co-occurred with WPS-2, Elsterales, KF-JG30-C25, Bradyrhizobium, and Ellin6067. Finally, Bryobacter co-occurred with Candidatus_Solibacter, Gemmatimonadaceae, and Bradyrhizobium.