Profiling of rhizosphere bacterial community associated with sugarcane and banana rotation system

Background Guangxi is the leading sugarcane-producing area in China. Due to the Panama disease outbreak in banana gardens, sugarcane and banana rotation was recommended. A field experiment with the newly released sugarcane cultivar Zhongzhe 1 (ZZ1) was conducted to understand the role of the sugarcane–banana rotation system in shaping the rhizosphere microbiota. Fields in the region possess characteristics of red laterite soil. Results Using Illumina HiSeq sequencing to analyze soil samples’ 16S rRNA V3-V4 region, the preceding banana rotation field had relatively greater bacterial diversity than the monoculture sugarcane field. Proteobacteria , Chloroflexi , Actinobacteria , and Acidobacteria were the dominant phyla, with distinct taxa enriched in each environment. However, the preceding sugarcane monoculture field enriched functional groups related to nitrogen fixation and cellulolysis. Network analysis highlighted contrasting network structures between sugarcane and banana rhizospheres, suggesting differential stability and susceptibility to environmental influences. Furthermore, correlations between soil properties and bacterial alpha-diversity underscored the influence of preceding crops on rhizosphere microbial communities. Conclusion This research enhances our understanding of crop rotation effects on soil microbial ecology and provides insights into optimizing agricultural practices for enhanced soil health and crop productivity. Future studies should explore the underlying mechanisms driving these interactions and evaluate the long-term impacts of crop rotation on soil microbial dynamics.


Introduction
Soil microorganisms are integral contributors to the functioning of soil ecosystems, contributing significantly to nutrient cycling, decomposition of organic matter, degradation of toxins, and resilience against biotic and abiotic stresses.Rhizosphere microorganisms, in particular, serve as valuable indicators of soil fertility and the effectiveness of land management practices, adeptly adjusting to various environmental and agricultural conditions [1].These microbes actively regulate the soil carbon and nitrogen cycles through various processes such as soil organic matter (SOM) decomposition, denitrification, nitrification, mineralization, immobilization, assimilation, and respiration [2,3].Furthermore, microbes may play a beneficial role in enhancing plant growth by facilitating water and nutrient uptake, preventing heavy metal accumulation, and inducing systemic resistance to pathogens [4][5][6].
Sugarcane is a crucial sugar and energy production crop, serving as a raw material source for many industrial products.Recent advancements in high-throughput sequencing have facilitated studies to gain insights into the composition and structure of the microbial communities inhabiting the sugarcane rhizosphere.Previous studies have delineated the core microbiome across several plant species, such as rice, wheat, tomato, citrus, herbal plants, and sugarcane.Plants have incredibly wide and complex interactions with microbes including mutualistic relationships upon which plants and microbiota depend to thrive in nature [7][8][9][10][11][12].Moreover, specific plant growth-promoting bacteria, such as Azospirillum, Bacillus, and Rhizobium, have been recognized with significant potential for utilization as bioinoculants for crop enhancement [13].Furthermore, the distinct rhizosphere and root bacteria profiles have been characterized among different sugarcane species and commercial cultivars [14][15][16][17].
Guangxi is a leading sugar-producing province in China, contributing more than 60% of the total national sugar production [18].In addition, it serves as one of the main banana-producing areas in China.However, the outbreak of Panama disease has rendered most banana orchards unsuitable for continuous cultivation.The adverse effects of continuous cropping of bananas on the composition of the soil microbial community harmed crop production, resulting in a decline in crop yield [4].The banana-sugarcane rotation has been available in Guangdong province in recent years.The incidence of Panama disease reduced from over 50% to 10% after rotation attributable to augmented levels of cultivable bacteria and decreased presence of Fusarium oxysporum and actinomycetes [19].
In recent years, some banana fields in Guangxi have undergone replanting with sugarcane.Consequently, there is a dire need to investigate the impact of different preceding crops on the rhizosphere bacterial community of sugarcane.Zhongzhe 1 (ZZ1), a newly released cultivar developed by our research team, is known for its high yield, sugar content, and resistance to smut was chosen for investigation.In the present milieu, high-throughput sequencing was used to analyze the diversity of bacterial communities in the rhizosphere of sugarcane cultivar ZZ1, planted in fields previously utilized for banana or sugarcane cultivation.The present study aimed to gain insights into the microbiota composition and putative functional roles within the context of the sugarcane monocropping and sugarcane-banana rotation system.

Experiment site and management
This experiment was conducted at the field station of Guangxi University in Fusui, China (22° 38′ 06″ N, 107° 54′ 15″ E).The preceding banana field designated for rotation intercropping had been planting bananas for over 6 years.Meanwhile, the sugarcane field had a history of continuous plantation for over a decade.Soil characteristics in the region included red laterite soil (Ultisols) with a pH of 5.5, 77.1 mg kg −1 available potassium, 83 mg kg −1 available phosphorous, 0.84 g kg −1 total nitrogen, 2.98 g kg −1 total phosphorous, and 19.47 g kg −1 organic matter.Uniform field management practices were maintained for both fields, with a planting density of 75,000 bud sets per hectare.The experiment area spanned over 40 hectares, about 2 km long and 1.5 km wide, divided into 15 plots (each plot of ~ 200 m × 100 m).

Soil samples collection
Soil sampling was performed on Dec 22, 2017 (6 months after planting).At each plot, a composite soil sample comprising ten homogeneously mixed subsamples was collected between two sugarcane lines.Soil adhering to the roots, considered rhizosphere soil, was collected by subjecting roots to vortexing with steel beads for 20 s, resulting in precipitation.Bulk soil samples were taken adjacent to the excavated sugarcane, approximately 15 cm from the stalk base and at a depth of 20 cm.Fourteen composite rhizospheres (PBR) and bulk (PBB) soil samples were collected from each preceding banana intercropping field.Fifteen composite rhizospheres (PSR) and bulk soil (PSB) samples were collected from 15 plots of preceding sugarcane monoculture fields (Figure S1).Each soil sample was ground and sieved through a 2-mm sieve.Then, each composite sample was divided into two parts: one stored at − 80℃ for DNA extraction and the other air-dried for subsequent chemical analysis.

Amplicon sequencing
The purified amplicons were pooled into equimolar proportions and subjected to paired-end sequencing (2 × 300) using the Illumina HiSeq 2500 platform.Detailed sample information and the generated sequences were archived in the NCBI Sequence Read Archive (SRA) database under accession number PRJNA730494.Subsequent processing involved the utilization of FLASH v.1.2.11 software to merge demultiplexed and quality-filtered raw FASTQ sequences [21].The criteria applied for this process were as follows: (i) truncation of reads at any position with an average score below 20 within a sliding window of 50 bp; (ii) removal of reads containing ambiguous bases, with perfect primer matches or having up to two nucleotides mismatches; (iii) merging of sequences with overlapping lengths exceeding 10 bp based on overlapping regions.Operational taxonomic units (OTUs) were clustered at 97% similarity using USEARCH (v10.0.) [22].The chimeric sequences were identified and eliminated using UCHIME (v 4.2.40).The taxonomic classification of each 16S sequence was analyzed against the 16S Ribosomal Database Project (RDP) database (http:// rdpcme.msu.edu/).

OTU network analysis from soil samples
Two OTU networks, PSR and PBR, were constructed based on relative abundance across 15 soil samples, with OTUs detected in 8 out of 15 samples used for network construction.Random matrix theory (RMT) determined the similarity threshold for network construction.A molecular ecological network analysis (MENA) pipeline was applied, and the network was visualized using Cytoscape v3.8.2 [27,28].Node was categorized as module hubs (Zi > 2.5), network hubs (Zi > 2.5 and Pi > 0.62), connector (Pi > 0.62), or peripheral equipment (Pi < 0.62 and Zi < 2.5), elucidating their roles within the network [28].

Soil chemical properties analysis
Soil pH was determined with a pH meter following extraction with 1 mol/L KCl extraction [29].Total nitrogen (TN) content was measured utilizing the automated nitrogen method (KMnO 4 − Fe − digestion in concentrated sulfuric acid, followed by titration with dilute sulfuric acid) [30].Soil organic matter (SOM) content was assayed using the potassium dichromate and sulfuric acid method [31].Available phosphorus (AP) was extracted using ammonium fluoride and subsequently measured through an ascorbic acid/molybdate assay [32].Available potassium (AK) was extracted with 1 mol Ammonium acetate and boiling nitric acid, with potassium concentration measured using a flame photometer (Shanghai Hengping Scientific Instruments Co., Ltd) [33].

Statistical analysis
The number of OTUs, bacterial community richness, alpha-diversity indices, and soil characteristics were assessed using one-way analysis of variance (ANOVA) with the LSD test via IBM SPSS statistics v.19 software.Wilcoxon rank-sum test was carried out to evaluate the relative abundance of dominant bacterial taxa, with statistically significant differences noted for p values (< 0.05) between groups.Pearson correlation heat map was generated using the same software, utilizing these p values to investigate the correlation between soil characteristics, bacterial alpha-diversity, and the relative abundance of dominant genera.Soil microbial community structure was evaluated by Anosim (analysis of similarities) at a significance level of 0.05, using the 'vegan' packages in R v.4.1.2software [34].
Similar to the preceding sugarcane field, the total relative abundances of Actinobacteria, Proteobacteria, Chloroflexi, and Acidobacteria, the four most abundant phyla, constituted 75.9% and 79.3% in preceding banana field rhizosphere (PBR) and bulk soil (PBB) samples, respectively.In PBR, 24 genera and in PBB, 29 genera exhibited abundances exceeding 1%, accounting for 45.8% and 51.5%, respectively, indicating a predominance of low-abundance genera and a potential lack of representatives among dominant genera.The top 5 dominant genera in PBR were uncultured_bacterium Gaiellales (3.49%) and uncultured_bacterium Micromonosporaceae (3.14%) from Actinobacteria, Sphingomonas (3.31%), uncultured_bacterium_c_Sub-group_6 (3.18%) from Acidobacteria, and uncultured_ bacterium_c_AD3 (2.84%).These findings indicate the consistent dominance of specific bacterial phyla in soils previously used for sugarcane monoculture or sugarcane-banana rotation cultivation, exerting a notable influence on the overall composition of microbial communities in these agricultural settings.

Diversity dynamics of soil bacterial communities across tested fields
The obtained sequences likely represented the majority of bacteria in each sample, as indicated by Good's coverage values exceeding 99% (ranging from 99.48% to 99.96%).No significant differences in the number of OTUs, Chao1, and Simpson index were observed between preceding sugarcane (PSR and PSB) and preceding banana (PBR and PBB) field samples.Conversely, although the Shannon index indicated no significant distinction between the rhizosphere (PSR) and bulk soil (PSB) samples derived from the preceding sugarcane field, it exhibited a significant contrast when compared to the rhizosphere (PBR) and bulk soil (PBB) samples from preceding banana field (Table 1).
PCoA provided insights into the community structure of each bacterial group.PERMANOVA results Fig. 1 Bacterial community composition at the A phylum and B genus levels in the rhizosphere and bulk soil of ZZ 1. PBR: rhizosphere soil from preceding banana fields; PBB: bulk soil from preceding banana fields; PSR: rhizosphere soil from monocultured sugarcane fields; PSB: bulk soil from monocultured sugarcane fields with p values above 0.05 indicated no significant differences between rhizosphere and bulk soil within the confines of the same plant field.However, a statistically significant difference was observed among the four groups (p = 0.001) (Fig. 2).PCoA of the rhizosphere (PSR) and bulk soil (PSB) from the preceding sugarcane monoculture revealed that these samples collectively accounted for only 46.6% of the total variation, with p value indicating a lack of significant disparity in their bacterial community compositions (Fig. 2A).Similarly, PCoA of the rhizosphere (PBR) and bulk soil (PBB) from preceding banana inter-crop accounted for only 40.83% of the total variation, with the associated p value suggesting an absence of significant heterogeneity in their bacterial community compositions (Fig. 2B).However, PCoA1 and PCoA2 axes accounted for 44.71% and 11.45% of the total variation in soil bacterial community structures, respectively, with p value symbolizing significant variation between preceding sugarcane (PSB, PSR) and banana (PBB, PBR) crop samples (Fig. 2C).PCoA ordination also illustrated significant distinctions in bacterial community structure among both plant fields.PSB and PSR plots cluster on the right, and PBB and PBR plots on the left.These outcomes revealed the variations in OTUs and diversity indices between monoculture sugarcane and preceding banana fields, with PCoA revealing distinct community structures.

Differentiated soil bacterial communities in preceding sugarcane and banana fields
LEfSe analysis revealed distinct taxonomic differences among the four bacterial groups (Fig. 3).The predominant bacterial taxa in the preceding sugarcane rhizosphere (PSR) encompassed Chloroflexi, Acidobacteria, WPS_2 phyla, and Frankiales order, while Firmicutes were more abundant in the preceding banana rhizosphere (PBR).Preceding sugarcane bulk soil (PSB) exhibited enrichment of Solibacterales and its family Solibacte-raceae_Subgroup_3.At the same time, the phyla Proteobacteria and Gemmatimonadetes, the class Subgroup_6, and the order Micromonosporales were significantly enriched in preceding banana bulk soil (PBB).Moreover, 158 genera were significantly different between PSR and PBR.
The Prokaryotic clades analysis showed greater abundances of communities related to aerobic chemoheterotrophy, cellulolysis, nitrogen fixation, and chemoheterotrophy in the rhizosphere of the preceding sugarcane field (PSR) compared to the preceding banana field (PBR).Conversely, nitrate reduction and fermentation-related communities were more prevalent in the PBR environment (Fig. 4).These outcomes conclude distinct taxonomic and functional differences between bacterial communities in preceding sugarcane and banana fields, emphasizing the enrichment of specific taxa and functions in each environment.

Differential network stability in rhizosphere bacterial communities
The network module separation and modularity analysis showed that most OTUs were positioned on the periphery, with intra-module connections predominating.Within the PSR environment, 11 nodes were identified as module hubs, while only 1 served as a connector (Fig. 5).These nodes originated predominantly from Firmicutes (four), Proteobacteria (three), Actinobacteria (two), Chloroflexi (one), and Acidobacteria (one).In contrast, the PBR environment featured five nodes as module hubs and two as connectors, with nodes primarily derived from Firmicutes (three) and one each from Fig. 2 Principal coordinate analysis (PCoA) of bacterial community compositions at OTUs level based on the Bray-Curtis dissimilarity matrix across sampling sites.A PCoA of bacterial communities between PSR and PSB groups.B PCoA of bacterial communities between PBR and PBB groups.C PCoA of bacterial communities among all field samples.Circles and ellipses are color-coded by groups and represent the standard deviation of axis scores from the group centroid.PBR: rhizosphere soil from preceding banana fields; PBB: bulk soil from preceding banana fields; PSR: rhizosphere soil from monocultured sugarcane fields; PSB: bulk soil from monocultured sugarcane fields Actinobacteria, Chloroflexi, Nitrospirae, and Thaumarchaeota (Fig. 5).
The correlations among OTU occurrences were calculated using the Pearson correlation coefficient (PCC) to gauge network topological properties.The rhizosphere of preceding sugarcane fields (PSR) exhibited a less modular yet highly complex structure.In contrast, the rhizosphere of preceding banana fields (PBR) displayed a more modular network with lower complexity (Fig. 6).PSR comprised 28 modules, with extensive interconnections among complex modules, especially between the most complex ones.Conversely, PBR comprised 35 modules with fewer interconnections between the most complex ones.Notably, only one module hub was identified in the most extensive module of PBR, suggesting a potentially more unstable network than PSR.The contrasting network structures between the rhizospheres of preceding sugarcane (PSR) and banana (PBR) fields suggest differential stability and susceptibility to external environmental influences in bacterial communities.

Rhizosphere soil properties influence bacterial community diversity
The soil properties exhibited slight variations across both fields (Table 2).In general, preceding banana field samples (PBR; PBB) exhibited higher soil composition parameters levels than preceding sugarcane field (PSR; PSB) samples.Relative to bulk soil samples, both rhizosphere samples (PSR; PBR) displayed lower soil pH, organic matter, total nitrogen (N), and available potassium (AK), while available phosphorus (AP) levels increased.Moreover, in the rhizosphere, PBR samples displayed increased soil pH, AP, and AK levels compared to PSR, with no significant difference in organic matter and total N content.
Pearson correlation of all soil samples showed a significant positive association between OTUs number (labeled as Feature), ACE, and Shannon indices of the bacterial communities with soil pH.Conversely, Shannon and Simpson indices exhibited significant negative correlations with total N content (Fig. 7A).Moreover, soil pH exhibited significant (p < 0.05) correlations with the relative abundance of dominant genera within the aforementioned abundant phyla Actinobacteria, Proteobacteria, Chloroflexi, and Acidobacteria (Fig. 7B).The genera uncultured_bacterium_c_Subgroup_6 and Gaiella, predominant in PBR samples, exhibited a notable positive correlation with soil pH and available potassium (AK) content.Conversely, genera uncultured_bacterium_c_ AD3 and Acidothermus, relatively abundant in PSR samples, were negatively associated with pH and AK levels.Furthermore, available phosphorus (AP) and total nitrogen (N) displayed negative correlations with two phyla as mentioned above, while both exhibited positive correlations with the remaining two.These results indicate associations between soil properties and bacterial alphadiversity, highlighting the impact of preceding crops on rhizosphere microbial communities and their interactions with soil characteristics.

Discussion
Rhizosphere microorganisms thrive near plant roots and are essential for plant development [35].Among these, bacteria are particularly active and prevalent in soil ecosystems.The richness and diversity of rhizosphere bacteria can significantly affect soil physiochemical properties Fig. 4 Metabolic and ecologically relevant putative functions of rhizosphere bacterial communities at OTUs level in ZZ1, indicating enhanced putative functionality in PSR compared to PBR.PSR: rhizosphere from monoculture sugarcane fields; PBR: rhizosphere from preceding banana fields Fig. 5 The analysis of network module separation and modularity calculation to identify critical bacteria (OTUs) in the rhizosphere of ZZ 1 from the previous sugarcane field (PSR) and previous banana field (PBR).The x-Zi represents within-module connectivity of specific modules or components (Zi) within the bacterial network, and the y-Pi represents among-module connectivity, indicating the connectivity between different modules or components (Pi) within the bacterial network and nutrient cycling.Notably, rhizosphere bacterial communities may exhibit distinct compositions across different geographical locations where the same plant species is grown [36].With the advancement of high-throughput sequencing technology, network analysis using microbial data has emerged as a widely favored method for processing big data efficiently.
Our results revealed that the predominant bacterial phyla in the field soil comprised Proteobacteria, Chloroflexi, Actinobacteria, Acidobacteria, Firmicutes, and Gemmatimonadetes.Notably, Proteobacteria, Acidobacteria, and Actinobacteria predominated in the rhizosphere of monoculture sugarcane fields, consistent with previous reports highlighting the prevalence of these  genera in continuous monoculture settings [37,38].In addition, Proteobacteria, Chloroflexi, Actinobacteria, and Acidobacteria were dominant in the rhizosphere soil of sugarcane-soybean intercropping fields, suggesting their typical dominance in sugarcane rhizosphere bacterial communities [39].Proteobacteria, owing to their diverse physiology, morphology, and metabolic capabilities, are ubiquitous in most plants and crucial for maintaining soil carbon and nitrogen cycles [40].Actinobacteria have been consistently linked to suppressing Fusarium wilt disease pathogens [41].Acidobacteria widely exist in the plant rhizospheres, contributing to polysaccharide degradation and plant growth enhancement [42].Chloroflexi may play essential roles in facilitating and detoxifying heavy metals for the plants in mining areas [43].Notably, the disparity between the rhizospheric bacterial communities of PSR and PBR was particularly evident in Acidobacteria, Chloroflexi, and Firmicutes, with Acidobacteria showing higher abundance in preceding sugarcane rhizosphere (PSR) (p = 0.003).These microbes may contribute to maintaining a healthy rhizome environment, thereby supporting plant health and sustainability.LefSe results also revealed enrichment of Acidobacteria in PSR, while FAPROTAX analysis indicated heightened cellulolytic activity in this environment.Acidobacteriales within Acidobacteria can degrade complex root exudates such as cellulose and lignin [44].Furthermore, PSR exhibited a relatively higher abundance of Bradyrhizobium, a genus specialized in nitrogen fixation and nodule formation [14].Candidatus_Nitrososphaera, predominant in PSR, potentially plays a crucial role in ammonia oxidation and denitrification, contributing to nitrogen cycle maintenance and atmospheric nitrogen removal [45,46].Consequently, the bacterial community of the preceding sugarcane rhizosphere (PSR) might be more efficient in nutrient utilization, thereby benefitting sugarcane growth and health.The richness (measured by OTUs number and ACE index) and diversity (Simpson index) of bacterial communities were positively correlated with soil pH, indicating greater diversity in neutral or weak alkaline conditions [47,48].Although no significant difference in Actinobacteria relative abundances between preceding sugarcane (PSR) and banana (PBR) rhizospheres was observed, Acidobacteria were relatively more abundant in PSR.Actinobacteria prefer weak alkaline soil, while Acidobacteria thrive in acidic soil.Despite a slight soil pH difference (0.4 units) between PSR and PBR, PSR exhibited more significant Acidobacteria enrichment.Moreover, the dominant genera correlated with soil pH and potassium content.Among them, the dominant genera of Acidobacteria showed a negative correlation with available potassium (AK) content, potentially due to potassium fertilizer inhibiting the proliferation of the acidophilic population in soil [49].The application of potassium fertilizer, aimed at improving soil nutrients, leads to the rapid increase in microbial species requiring abundant nutrients, consequently reducing the abundance of microbial species with lower nutrient requirements [50].These correlation analyses between potassium and dominant genera provide a reference for optimizing soil microbial composition through adjustments in potassium fertilizer application in banana fields.
Twenty-eight modules were identified within the constructed microbial network of the preceding sugarcane rhizosphere (PSR), with the top four modules exhibiting the most significant number of nodes and connections.Six modules harbored 12 key OTUs, predominantly Firmicutes, Proteobacteria, and Actinobacteria.In contrast, the rhizosphere microbial network of the preceding banana field (PBR) comprised 35 modules, with the two most prominent modules having the fewer connections.Four modules obtained eight key OTUs, mainly from the phylum Firmicutes.The PSR network exhibited greater complexity and less modularity, indicating a highly interconnected bacterial community structure.Such a network structure suggests potential implications for microbial community diversity and resistance to pathogen invasion [51].Communities with higher diversity levels may exhibit increased resistance to pathogen invasion [52] due to enhanced species interactions and competition for niches.In addition, modular microbial communities with shorter path lengths might react more rapidly to environmental disturbances [53].
Comparative analysis revealed that the preceding sugarcane rhizosphere (PSR) group displayed greater average connectedness, aggregation coefficients, and shorter path lengths than the preceding banana rhizosphere (PBR) group.Based on these results, we hypothesize that the bacterial network of the PSR group may be more resistant to pathogen invasion.In contrast, the bacterial network of the PBR group is more susceptible to environmental interference.Further studies are required to verify this hypothesis and assess its implications for disease incidence in ZZ1 across different preceding crop fields.

Conclusion
This study delved into the repercussions of sugarcanebanana rotation on bacterial community profiles of sugarcane cultivar ZZ1, revealing distinct taxonomic and functional differences between preceding sugarcane and banana fields.Our results underscore the influence of crop rotation on soil microbial dynamics, leading to shifts in bacterial richness and community structure.Notably, a robust and analogous relationship was observed within the rhizosphere of ZZ1 in preceding sugarcane fields compared to preceding banana fields.Despite a decline in bacterial population density within the rhizosphere, the preceding sugarcane environment exhibited significantly higher abundance and functional diversity, particularly evident in critical ecological processes such as cellulolysis and nitrogen fixation, indicative of its positive influence on plant health.Moreover, the intricate interplay among rhizosphere bacterial populations was more intricate in preceding sugarcane fields, highlighting the pivotal role of sugarcane cultivation in shaping rhizosphere microbial communities.However, it is imperative to recognize the multifaceted nature of soil and rhizosphere microbial dynamics, which are influenced by various factors, including soil properties, enzymatic activities, temperature fluctuations, and plant genotype variations.Hence, comprehensive investigations are warranted to elucidate the holistic effects of crop rotation on sugarcane rhizosphere activities while considering the intricate interplay with other critical environmental parameters.

Fig. 3
Fig. 3 Cladogram generated from LEfSe analysis showing the taxonomic levels represented by rings with phyla in the innermost ring and genera in the outermost ring.Circle size reflects taxon abundance, color indicates sample type.Blocks on the right, colored in red, green, orange, and blue, signify significantly higher relative abundances in different sample types: PBR (orange) representing rhizosphere soil from preceding banana fields, PBB (blue) representing bulk soil from preceding banana fields, PSR (red) representing rhizosphere soil from monocultured sugarcane fields, and PSB (yellow) representing bulk soil from monocultured sugarcane fields.Darker shades of rings indicate more substantial similarities within taxa, while lighter shades highlight differences between taxa.The various taxonomic groups are annotated with letters (e.g., a, b, c) representing different orders, families, and genera, such as Acidobacteriales, Solibacterales, Rhizobiales, and Clostridiales

Fig. 6
Fig. 6 Co-occurrence networks of rhizobacterial communities in A preceding sugarcane and B preceding banana fields.Node sizes correspond to OTUs relative abundances, colored by phylum.Pink links indicate positive covariation and blue ones indicate negative covariation.Different node shapes represent distinct keystone species.Red arrows indicate key OTUs identified by network analysis

Fig. 7
Fig. 7 Pearson's correlation coefficients between bacterial alpha-diversity indices and A dominant genera; B soil properties at OTUs level.Total N: total nitrogen; P: available phosphorus; AK: available potassium.The color and asterisk represent the R-value and significance level (* p < 0.05, ** p < 0.01, *** p < 0.001)

Table 1
Alpha-diversity indices based on different soil samplesDifferent letters indicate significant differences (ANOVA, p < 0.05, Waller-Duncan analysis) among soil samples PBR: rhizosphere soil from preceding banana fields; PBB: bulk soil from preceding banana fields; PSR: rhizosphere soil from monocultured sugarcane fields; PSB: bulk soil from monocultured sugarcane fields

Table 2
Relative difference in soil composition of rhizosphere and bulk soil of preceding banana field and sugarcane monocrop field samples PBR: rhizosphere soil from preceding banana fields; PBB: bulk soil from preceding banana fields; PSR: rhizosphere soil from monocultured sugarcane fields; PSB: bulk soil from monocultured sugarcane fields * Significance at the level of p < 0.05; ** great significance at the level of p < 0.01