Skip to main content
NCBI home page
Plant Diversity logoLink to Plant Diversity
. 2026 May 30;48(5):997–1010. doi: 10.1016/j.pld.2026.05.013

Leaf nutrients regulate phyllosphere microbial dynamics in rubber trees

Jiao Sun a,b, Yaqing Wei b, Yanjun Du a,⁎, Guoyu Lan b,c,⁎
PMCID: PMC13615783  PMID: 42801097

Abstract

Mechanisms underlying the assembly and succession of phyllosphere microbiota in rubber plantations are crucial for understanding ecosystem health and functionality, yet poorly understood. In this study, high-throughput sequencing was used to analyze the diversity, assembly mechanisms, and potential sources of phyllosphere microbial communities at developmental stages. Key findings include: (1) Epiphytic and endophytic bacterial diversity peaked at the budding stage, then decreased and subsequently showed a gradual increase again toward the senescence stage, primarily driven by leaf nitrogen (N), phosphorus (P), and potassium (K). However, epiphytic and endophytic fungal diversity was significantly regulated by leaf N and P. (2) Assembly processes exhibited niche differentiation: the assembly of epiphytic bacteria during the color-changing and stable stages was dominated by stochastic processes, while deterministic regulation prevailed in other stages and in endophytic communities. Both assembly and network complexity were driven by leaf N and K dynamics. Fungal network complexity and deterministic regulation, in both epiphytic and endophytic niches, were promoted by leaf N and P during early stages but gradually weakened thereafter. (3) Active material exchange occurred between epiphytic and endophytic niches, and below-ground niches contributed more significantly to the fungal community. Overall, our results underscores that leaf nutrients modulate the diversity, assembly, and resultant compositional landscape of phyllosphere microorganisms at different developmental stages. These insights contribute to deepening the fundamental understanding of plant-microbial symbiosis and inform strategies for the cultivation of rubber trees.

Keywords: Phyllosphere microorganisms, Leaf development, Rubber tree, Microbial dynamics

Highlights

  • •

    Bacterial and fungal diversity fluctuations exhibit distinct patterns influenced by nutrient availability.

  • •

    Community assembly demonstrates niche differentiation, primarily governed by foliar nitrogen and potassium levels.

  • •

    Active material exchange occurs between epiphytic and endophytic; below-ground niches contributed more to the fungal community.

1. Introduction

Phyllosphere microorganisms (Lindow and Leveau, 2002; Lindow and Brandl, 2003), a collective term encompassing all microorganisms colonizing epiphytic and endophytic microbial communities of plant leaves (Vorholt, 2012), fulfilling critical functions in regulating plant growth, pathogen resistance, and nutrient cycling (Barbour and Martiny, 2024). Hevea brasiliensis (rubber tree), as the sole botanical source of commercially viable natural rubber (Carnier and Martinelli, 2005), constitutes a cornerstone of the economic frameworks of tropical regions (Wei et al., 2025). Rubber tree leaves undergo six developmental stages: Budding, Bronze, Color-changing, Light-green, Stable, and Defoliation (Li et al., 2016). The composition of the rubber tree phyllosphere microbiome is influenced by multiple factors, including host plant species, leaf age (Wang et al., 2024), and environmental conditions (Laforest-Lapointe et al., 2017). Plant developmental stages explain substantial variation in endophytic bacterial and fungal communities, selectively shaping the composition and function of their associated microbial communities (Yang et al., 2025). However, compared to the well-studied rhizosphere microbiome, the assembly rules and underlying driving mechanisms of the rubber tree phyllosphere microbial community remain poorly understood. Elucidating the assembly mechanisms of phyllosphere microbial communities is a core issue in ecology (Wang et al., 2023).

The assembly mechanisms of microbial communities remain a complex and critical area of research (Li et al., 2020), encompassing the dynamic changes in microbial populations across different environments (Evans et al., 2020), interspecific interactions, and the formation and evolution of community composition (Khandoori et al., 2024). Classical ecological theories offer two distinct perspectives for exploring this complexity. Neutral theory posits that stochastic ecological drift of species drives community dynamics, leading to random community characteristics (Kimura, 1991; Hao et al., 2016). Some studies suggest that models based on neutral theory can effectively simulate species distribution in natural communities (Ke-chang et al., 2009). In contrast, niche theory emphasizes that deterministic processes, such as environmental filtering and interspecific interactions, collectively shape community structure (Peipoch et al., 2019; Chu et al., 2021), although its applicability may be limited under specific circumstances (Xu et al., 2016). However, the relative importance of stochastic and deterministic processes in community assembly is non-negligible (Jeraldo et al., 2012), and this relative importance varies dynamically with habitat type, environmental conditions, and spatiotemporal scales (Kalyuzhny et al., 2014; Wang et al., 2023).

Therefore, to elucidate the conditions influencing the relative importance of these processes, we selected the dynamic ecological environment of the rubber tree phyllosphere. As rubber tree development progresses, the allocation of leaf nutrient factors changes accordingly. Factors such as senescence, stress conditions (e.g., cold or drought damage), and the health status of plant (Li et al., 2024) influence the reallocation of nutrients among different plant organs (Masclaux et al., 2000; Millard and Grelet, 2010). Studies have found that leaf mineral element content and nitrogen fertilizer application rates can significantly influence the structure and diversity of phylloplane microbial communities (Muneer et al., 2025). Increased nitrogen fertilizer application often leads to a decline in bacterial community alpha diversity, while the response of fungal community diversity varies depending on the plant species (Darlison et al., 2019).

Existing research often employs static sampling designs, leaving a significant gap in our understanding of how plants utilize the developmental process to regulate microbial colonization and assembly during the dynamic progression from leaf development to senescence. Niche differentiation has been identified as a key driver of microbial diversity and species interactions in river floodplains (Peipoch et al., 2019) and acidic reservoirs (Soria-Píriz et al., 2024). Research indicates that endophytic bacterial communities are more susceptible to leaf senescence than their epiphytic counterparts (Lan et al., 2024). The epiphytic and endophytic compartments of the leaf constitute distinct microhabitats, offering a natural comparison for examining how habitat heterogeneity shapes assembly processes. However, the coordinated changes between these two microhabitats throughout leaf development remain systematically unexplored. More importantly, bacteria and fungi, as the two major microbial groups in the phyllosphere, may exhibit fundamentally different responses to the same developmental stage changes due to their contrasting life history strategies. This divergence has not yet been revealed in rubber trees.

Based on the current research landscape, this study proposes three hypotheses: (1) As the leaves of rubber trees initiate growth from the budding stage, their surfaces are essentially devoid of microbial colonization. Throughout subsequent developmental phases, the microbial taxonomic richness within the phyllosphere exhibit a progressive augmentation. (2) The complexity of phyllosphere microbial networks was low during the initial stages of development and gradually increased over time. Community assembly was primarily driven by stochastic processes in the early stages, while deterministic processes progressively strengthened and became dominant during the stable stages. (3) As leaf development progressed, contributions from soil and epiphytic microorganisms gradually increased and reached a relatively balanced state during the stable stage.

2. Materials and methods

2.1. Sampling site and experimental design

The sampling site for this experiment was located in Danzhou City, Hainan Province. Danzhou is situated in the northwest of Hainan Island, characterized predominantly by hilly terrain interspersed with plains. The region has distinct dry and wet seasons and exhibits a typical monsoon climate, with synchronized heat and precipitation, abundant sunshine, and ample rainfall. The specific sampling location was at the Third Team of the experimental station operated by the Chinese Academy of Tropical Agricultural Sciences (CATAS) in Danzhou City, situated in the northwestern region of Hainan Island (19°32′N, 109°28′E, elevation approximately 114 m). The terrain at the site is flat, and the rubber trees planted there were of the variety “Reyan 7-33-97”, known for its strong resistance, distinct characteristics, and suitability for sampling. The sampled trees in the mature plantation were planted in 2001.

At the Third Team experimental site, five sampling points were established. Triplicate sampling was conducted in March 2022, July 2022, and March 2023 to cover six developmental stages of rubber tree leaves: Budding, Bronze, Color-changing, Light-green, Stable, and Defoliation (Fig. 1B). At each sampling point, three rubber trees were selected, spaced approximately 100 m apart. From each standard tree, leaves were collected from four directions (east, west, south, north) at a distance of about 2 m from the trunk center and approximately 12 m above ground. Approximately ten leaves were collected per direction. Leaves from each direction were combined into a single composite sample. From each sampling point, about 50 leaves representing different developmental stages were collected from the three trees, placed in sterile sealed bags, and transported to the laboratory under low-temperature conditions for subsequent nutrient content determination and microbial sequencing.

Fig. 1.

Fig. 1

Open in a new tab

Phyllosphere changes at different developmental stages in a rubber tree plantation. (A) Variations in nutrient factors. (B) Leaf morphological changes. Post hoc Tukey's tests were used to determine which specific group means were significantly different from each other after finding a significant effect in the ANOVA. Note: identical letters indicate no significant difference (p > 0.05), while different letters indicate significant differences (p < 0.05). Bud: Budding, Br: Bronze, Cc: Color-changing, Lg: Light-green, St: Stable, De: Defoliation.

To obtain epiphytic microbial community samples, approximately 15–20 g of leaves were placed in sterile glass bottles containing sterile buffer; 10–15 g of roots and branches were placed in sterile centrifuge tubes containing PBS solution. Samples were subjected to 1 min of sonication followed by shaking on an orbital shaker, and this extraction procedure was carried out in triplicate to obtain epiphytic microbial samples from the leaves, branches, and roots. The resulting suspension was passed through a 0.22 μm pore-size filter, after which the membrane containing the microbial biomass was transferred into an aseptic microcentrifuge tube. To obtain endophytic microbial community samples, after epiphytic microorganism extraction, leaves, roots, and branches were rinsed with sterile water, subjected to surface sterilization, washed five times with sterile water (Bodenhausen et al., 2013; Ruiz-Pérez et al., 2016). Finally stored in sterile centrifuge tubes at low temperature (−80 °C) for subsequent analysis.

In this study, N content in leaves and branches was determined using the potassium persulfate oxidation method, P content was measured via the molybdenum antimony anti-absorption colorimetric method, K content was analyzed using a flame photometer, and C content was assessed by the potassium dichromate oxidation method.

2.2. DNA extraction and PCR amplification

Microbial community genomic DNA was isolated using the E.Z.N.A.® Soil DNA Kit (Omega Bio-tek, Norcross, GA, USA) following the manufacturer's recommended protocol. Quality assessment of the extracted genomic DNA was performed through 1% agarose gel electrophoresis, while quantification of DNA concentration and evaluation of purity were carried out with a NanoDrop 2000 spectrophotometer (Thermo Scientific, USA). Amplification of the fungal ITS1 region was performed with the primer pair ITS1F (5′-CTTGGTCATTTAGAGGAAGTAA-3′) and ITS2R (5′-GCTGCGTTCTTCATCGATGC-3′) (Adams et al., 2013). For bacterial amplification, the primer pair 515FmodF (5′-GTGYCAGCMGCCGCGGTAA-3′) and 806RmodR (5′-GGACTACNVGGGTWTCTAAT-3′) was used (Sampson et al., 2016; Walters et al., 2016).

The PCR protocol for 16S comprised an initial denaturation step (95 °C, 3 min), succeeded by 29 cycles each consisting of denaturation (95 °C, 30 s), primer annealing (53 °C, 30 s), and strand extension (72 °C, 45 s), concluding with a final elongation phase (72 °C, 10 min). The PCR protocol for ITS comprised an initial denaturation step (95 °C, 3 min), succeeded by 35 cycles each consisting of denaturation (95 °C, 30 s), primer annealing (55 °C, 30 s), and strand extension (72 °C, 45 s), concluding with a final elongation phase (72 °C, 10 min). Amplifications were conducted with the 10× TaKaRa rTaq DNA Polymerase reaction system (Wei et al., 2022). Target amplification products were separated by electrophoresis on a 2% agarose gel, excised, and subjected to purification with the AxyPrep DNA Gel Extraction Kit (Axygen Biosciences, Union City, CA, USA). DNA concentration was subsequently determined using the QuantiFluor™-ST fluorometric quantification platform (Promega, USA). Following gel electrophoresis, PCR products were purified employing the QIAquick gel extraction kit (Qiagen). For subsequent ASV (Amplicon Sequence Variant) analysis, sequence processing was performed using the DADA2 denoising pipeline (2024 version, https://qiime2.org). Reads were quality-trimmed by removing terminal regions where the Phred quality score fell at or below 20. Paired-end reads were then assembled into contigs of roughly 300 base pairs. Final curation involved quality-based filtering (maximum expected error threshold set at 0.2) and the exclusion of singleton sequences, executed with USEARCH software v.10 (Edgar, 2010). The purified amplicons underwent paired-end sequencing (2× 250 bp) using the Illumina MiSeq platform following the standard operating procedures provided by the manufacturer.

2.3. Data analysis

The raw fastq files underwent demultiplexing and quality filtering using QIIME2 (Bolyen et al., 2019). Taxonomic assignment for each variant was conducted employing the RDP Classifier tool (release 11.5) based on the 16S ribosomal RNA gene reference database GreenGenes (v.135, http://greengenes.secondgenome.com/) and ITS database UNITE (v.8.0, https://unite.ut.ee/) using a confidence threshold of 0.7 (Abdelfattah et al., 2021). Non-bacterial ASVs (e.g., assigned to Mitochondria or Chloroplast) were subsequently removed (Laforest-Lapointe et al., 2017). The paired-end sequencing yielded 11,405,781 high-quality bacterial reads and 11,675,283 fungal reads. Following quality control and denoising, the processed sequencing reads were grouped into distinct amplicon sequence variants (ASVs) based on 100% sequence identity, resulting in 39,311 bacterial and 66,818 fungal ASVs, respectively. To minimize the impact of sequencing depth on subsequent analyses of α- and β-diversity, datasets were rarefied to the minimum sequencing depth of each kingdom: all 31,512 sequences per sample for bacteria and 16,868 sequences per sample for fungi.

α-diversity of the phyllosphere microbial communities was assessed using metrics such as the Chao1 index, calculated from the identified ASVs per sample with the vegan package in R v.4.4.2 (Oksanen et al., 2013). Differences in α-diversity indices among different groups were evaluated for significance using the Wilcoxon rank-sum test implemented via the wilcox.test function in the stats package of R v.4.4.2 (R Core Team, 2024), with multiple comparisons adjusted via the false discovery rate (FDR) correction. Moreover, in subsequent data analyses, we have applied FDR (Benjamini-Hochberg) adjustment to all p-values to ensure the robustness of the results.

Principal coordinate analysis (PCoA) based on Bray–Curtis distance was performed using the vegan package in R v.4.4.2 (Oksanen et al., 2013) to examine the similarity of microbial community structures (β-diversity) among samples. The statistical significance of microbial community structural variation across developmental stages was evaluated through permutational multivariate analysis of variance (PERMANOVA), implemented with the adonis2 function in the vegan package (R v.4.4.2) (Oksanen et al., 2013).

To investigate the variations in microbial community co-occurrence patterns across different developmental stages, network analysis was conducted on leaves from six distinct developmental stages of the rubber tree based on Spearman correlation. To minimize potential spurious correlations arising from rare taxa, only abundant bacterial and fungal taxa (the top 1000 abundant ASVs) were considered. For network construction, robust correlations were filtered using widely applied thresholds: an absolute Spearman correlation coefficient (ρ) > 0.6 and a false discovery rate (FDR)-corrected p < 0.05 (Jiao et al., 2022). Modularity was then estimated, and the networks and modules were visualized using the MetaNet package in R v.4.4.2 (Peng, 2025).

The relative contributions of deterministic versus stochastic assembly processes in phyllosphere microbial communities were quantified using the Modified Stochasticity Ratio (MST) metric, MST values below 50% indicate more deterministic assembly, while values above 50% indicate more stochastic assembly (Ning et al., 2019; Lan et al., 2023). The MST values was performed with the NST software package implemented in R v.4.4.2 (Peng, 2025), and significant differences in MST across the six developmental stages were tested using the nst.panova function. Subsequently, the MST indices of the microbial communities were visualized using ggplot2 (Wickham, 2016). The fast expectation-maximization microbial source tracking method (FEAST) was implemented via the FEAST package in R v.4.3.1 (Su et al., 2021) to track changes in the proportional contributions of different source communities to the phyllosphere microbial community during leaf development.

Linear regression was performed using the vegan package (Oksanen et al., 2013) in R v.4.4.2. The Mantel test was applied via the mantel_test function in the linkET package (Huang, 2021) to examine the relationships between the composition, diversity, and network complexity of the phyllosphere microbial community and leaf nutrient elements (including N, P, K, and C) across different leaf developmental stages. Bacterial and fungal network complexity was represented by the average degree. Environmental-microbial correlations were graphically displayed by converting the corresponding numerical matrices into visual formats.

3. Results

3.1. Changes in nutrient factors

Analysis of the nutrient contents in rubber tree leaves across different developmental stages revealed the following: N, P, and K contents were higher during the bronze and light-green stages, then gradually declined as leaf development progressed, reaching their lowest concentrations at the defoliation stage. C content was lowest during the initial developmental stages and gradually increased with leaf development, stabilizing from the color-changing stage onward (Fig. 1A).

3.2. Differentiation of phyllosphere microbial communities

Bacterial community diversity showed a transient decrease from the budding to bronze stage, followed by a gradual increase, with Chao1 richness reaching its peak during the defoliation stage (p < 0.001; Fig. 2A). For fungal community diversity, Chao1 richness was significantly higher during the budding stage compared to other periods (p < 0.05), reached its lowest point at the color-changing stage, and subsequently increased gradually (Fig. 2B). Mantel test results revealed that α-diversity of phyllosphere bacteria increased progressively with the decline in nutrient factors N/P/K (Fig. 2C). Fungal diversity increased during the budding and defoliation stages as nutrient factors N/P decreased (Fig. 2D).

Fig. 2.

Fig. 2

Open in a new tab

Microbial diversity across different developmental stages of rubber tree. (A, B) Chao1 index (α-diversity) of bacterial and fungal communities. Significance levels: ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001. (C, D) Environmental drivers of bacterial and fungal community diversity. The upper-right section displays pairwise comparisons of environmental factors, with color gradients indicating Spearman's correlation coefficients. Bud: Budding, Br: Bronze, Cc: Color-changing, Lg: Light-green, St: Stable, De: Defoliation.

To further investigate the drivers underlying changes in microbial diversity, we analyzed community composition. Principal coordinate analysis (PCoA) combined with permutational multivariate analysis of variance (PERMANOVA), both derived from Bray–Curtis dissimilarity matrices, revealed that except during the defoliation stage, bacterial communities clustered primarily by niche (epiphytic and endophytic), indicating highly significant differences in β-diversity between epiphytic and endophytic bacterial communities (p < 0.001; Fig. 4A). However, phyllosphere fungal communities showed the most pronounced differentiation only during the budding (R2 = 0.1, p < 0.001) and defoliation stages (R2 = 0.17, p < 0.001; Fig. 4B). These findings suggest that niche exerts a significant influence on bacterial communities assembly than on fungal communities, while fungal communities are more responsive to developmental stage.

Fig. 4.

Fig. 4

Open in a new tab

Principal coordinates analysis (PCoA) of taxonomic similarity based on Bray–Curtis distances community composition. Significance levels: ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001. Bud: Budding, Br: Bronze, Cc: Color-changing, Lg: Light-green, St: Stable, De: Defoliation.

Furthermore, γ-proteobacteria and α-proteobacteria were identified as dominant bacterial classes (Figs. 3A, B and S1) and exhibited the most pronounced responses to developmental stage. Within the fungal community, Dothideomycetes and Sordariomycetes were the dominant classes that displayed greater shifts in abundance in response to leaf development (Figs. 3C, D and S1).

Fig. 3.

Fig. 3

Open in a new tab

Taxonomic composition of bacteria and fungi at different developmental stages (at the class level). (A, C) Bar chart of the relative abundance of bacterial phyla and fungal phyla. (B, D) Line chart of the relative abundance of dominant bacteria and fungi. The classes marked with green and red asterisks indicate significant differences in their abundance at different developmental stages on leaf epiphytic and endophytic niches. Significance levels: ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001. Bud: Budding, Br: Bronze, Cc: Color-changing, Lg: Light-green, St: Stable, De: Defoliation.

3.3. Dynamic changes in the phyllosphere microbial community network

Co-occurrence patterns of the phyllosphere microbiota were evaluated across different developmental stages. The topological characteristics of network structures differed significantly between bacterial and fungal communities. For bacterial networks, the number of edges and average degree in the network gradually increased over the developmental stages, indicating progressively enhanced connectivity and complexity (Fig. 6C). The number of modules followed a “U-shaped” trend from budding to stable stages (6-3-5), reflecting an initial decline followed by a rise in network specialization. In terms of taxonomic composition, Proteobacteria and Actinobacteria constituted the highest proportions of participating taxa indicating that senescing leaves may provide ecological niches for rare or potentially pathogenic bacteria (Fig. 5A).

Fig. 6.

Fig. 6

Open in a new tab

Environmental drivers of bacterial (A) and fungal (B) community network complexity. The upper right corner shows pairwise comparisons of environmental factors, with color gradients indicating Spearman's rank correlation coefficients. (C, D) Comparison of average degree across development stages. Post hoc Tukey's tests were used to determine which specific group means were significantly different from each other after finding a significant effect in the ANOVA. Note: identical letters indicate no significant difference (p > 0.05), while different letters indicate significant differences (p < 0.05). Bud: Budding, Br: Bronze, Cc: Color-changing, Lg: Light-green, St: Stable, De: Defoliation.

Fig. 5.

Fig. 5

Open in a new tab

Analysis of phyllosphere bacterial (A) and fungal (B) community network structures across different developmental stages.

For fungal networks, the number of edges and average degree initially increased during early to mid-development before declining in the final stage, indicating that network complexity and network connectivity peaked at the bronze stage and subsequently diminished (Fig. 6D). The proportion of negative correlation edges was reached a peak during the stable stage (18.1%), while remaining at 0.8% during defoliation. This pattern indicates that fungal community networks were most stable and exhibited greater persistence during the stable stage. Compared to bacterial modularity, fungal modular specialization commenced earlier and maintained a higher number of modules even during leaf senescence. This suggests that fungal communities possess a stronger capacity for micro-niche differentiation and greater ecological persistence. Furthermore, Basidiomycota and Ascomycota nodes exhibited a slight increase during defoliation, which may be associated with later-stage decomposition processes (Fig. 5B).

Overall, the complexity of fungal networks increased in the early and middle stages but decreased in the later stages, contrasting sharply with the gradual increase in complexity of bacterial networks. However, fungal networks exhibited stronger specialization and more persistent stability. Further analysis indicated that N/K were the most important predictors of bacterial network complexity (p < 0.01), with bacterial network complexity increasing as nutrient factors N/K decreased (Fig. 6A). In contrast, N/P were the most important predictors of fungal network complexity (p < 0.05; Fig. 6B).

3.4. Stochasticity/determinism analysis of phyllosphere microbial communities

Epiphytic bacterial communities were predominantly shaped by stochastic processes during the color-changing and stable stages, with the highest stochasticity observed during the color-changing stage (MST = 53.71%). In contrast, deterministic processes dominated during the other stages, with the strongest deterministic influence on epiphytic bacterial communities occurring during the budding stage (MST = 19.85%). Throughout leaf development, endophytic bacterial communities were consistently dominated by deterministic processes, with the strongest deterministic influence during the bronze stage and the highest stochasticity during the defoliation stage (MST = 48.31%; Fig. 7A). For fungal communities, both epiphytic and endophytic communities were primarily driven by deterministic processes across all developmental stages (Fig. 7B). Further analysis revealed that the modified stochasticity ratio (MST) of bacterial communities was significantly correlated with N content (p < 0.05; Fig. 7C), whereas N/P contents were the main factors influencing the MST of fungal communities (p < 0.01; Fig. 7D).

Fig. 7.

Fig. 7

Open in a new tab

Variations in bacterial and fungal community assembly patterns across different developmental stages. (A, B) Modified stochasticity ratio (MST) of bacterial and fungal communities (MST < 50% indicates deterministic dominance in community assembly, MST > 50% indicates stochastic dominance in community assembly). (C, D) Relationships between bacterial and fungal assembly processes and nutrient factors.

3.5. Source tracking of phyllosphere microorganisms

Epiphytic and endophytic leaf compartments serve as both sources and sinks for microbial communities, while microbial exchange with below-ground niches is notably limited. Source tracking analysis via FEAST revealed that epiphytic bacteria and fungi primarily originated from the leaf endosphere and branches. Contributions from the leaf endosphere were particularly high during the bronze stage (43.90% and 41.49%) and the stable stage (52.45% and 52.94%). Contributions from branch surfaces were highest at the budding stage (12.38% and 11.70%) and subsequently declined. Contributions from the branch endospheres peaked during the defoliation stage (13.33% and 12.51%). The combined contribution from all below-ground compartments (soil, rhizosphere, rhizoplane, and root endosphere) remained below 8.00% across all developmental stages, reaching a relatively higher value only during the defoliation stage (7.79% and 7.94%). Endophytic bacteria and fungi likewise originated predominantly from phylloplane and branch surface. Contributions from epiphytic sources peaked during the bronze stage (50.69% and 49.00%) and the stable stage (42.00% and 42.03%). Contributions from branch surfaces were relatively high during the color-changing stage (8.03% and 8.30%) and the defoliation stage (6.38% and 6.48%). For bacteria, the contribution from the branch endosphere was highest during the budding stage (8.53%), while for fungi, it peaked during the bronze stage (8.44%), after which contributions from this source gradually decreased. Meanwhile, the proportion originating from below-ground compartments increased after the color-changing stage, reaching its maximum during defoliation (3.13% and 2.98%). Microbial exchange with below-ground compartments was minimal during the stable stage (Fig. 8A and C). However, the contribution of below-ground niches was greater for fungi than for bacteria (Table S1).

Fig. 8.

Fig. 8

Open in a new tab

Microbial source tracking analysis by FEAST across different developmental stages. Bud: Budding, Br: Bronze, Cc: Color-changing, Lg: Light-green, St: Stable, De: Defoliation. En: Endophytic, Ep: Epiphytic. P: Phylloplane, PE: Leaf endosphere, Branch: Branch surface, Branch_Endo: Branch endosphere, Root_Endo: Root endosphere.

4. Discussion

4.1. Nutrient-mediated microbial succession

The composition of bacterial and fungal communities undergoes dynamic changes with growth stages, with dominant microbial communities showing a successional trend across different developmental phases (Gao et al., 2023). Furthermore, studies indicate that leaf growth and senescence stages influence the composition and metabolic activities of the microbiome, with more pronounced dynamic shifts particularly evident during the defoliation period (Voříšková and Baldrian, 2013; Purahong et al., 2016). In rubber trees, the relative abundances of dominant bacterial groups, including α-proteobacteria and γ-proteobacteria, and dominant fungal groups, including Dothideomycetes and Sordariomycetes, exhibit significant and distinct fluctuation patterns across various leaf developmental stages. These microbial communities collectively undergo structural reorganization during leaf defoliation, a phenomenon that has been validated and elucidated through multiple studies (Lan et al., 2024).

Previous research has shown that many fungi within Dothideomycetes possess saprophytic capabilities (Weatherhead et al., 2022) and are important parts in maintaining host plant health (Mendes et al., 2011). The relative abundance of Dothideomycetes in the leaf endophytic peaked during the stable stage, showing a significant negative correlation with leaf N/P/K contents. In contrast, Sordariomycetes exhibited a significant positive correlation only with leaf N content. These findings indicate that during the stable stage, Dothideomycetes fungi with decomposing potential are enriched within the leaf interior, likely due to environmental filtering driven by changes in the internal leaf micro-environment. We hypothesize that in later developmental stages, shifts in phylloplane conditions may favor the colonization of saprophytic Dothideomycetes fungi from the surrounding environment. This temporal succession of fungal communities reflects passive microbial responses to the changing micro-environmental conditions during leaf development and senescence. Dothideomycetes fungi from the environment to colonize phylloplane to facilitate decomposition. This phenomenon may reflect a life-history strategy adopted by rubber tree leaves during growth and development. During the stable stage, changes in the phyllosphere micro-environment may create conditions favorable for the colonization of fungi with decomposing capabilities, potentially setting the stage for subsequent leaf abscission and nutrient cycling (Xiong et al., 2021a). This phenomenon likely reflects a passive shift in fungal community composition as leaves age, driven by an increasing emphasis on decomposition. Additionally, as host leaves approach senescence and their immunity declines, they become more susceptible to colonization by saprophytic fungi (Xiong et al., 2021b). Concurrently, the relative abundance of Sordariomycetes also significantly increased during the defoliation stage. The observed enrichment of Sordariomycetes may be attributed to its endophytic nature, which could enhance host nutrient use efficiency through endophytic symbiosis and mycoparasitic competition mechanisms (Vujanovic, 2021), while also potentially contributing to ecosystem disease resistance.

The observed pattern of a gradual increase in the relative abundance of α-proteobacteria and a decrease in γ-proteobacteria may be attributed to their distinct nutrient utilization strategies. Research findings indicate that α-proteobacteria increased when leaf N/P/K contents were low, whereas γ-proteobacteria rose as these nutrient contents declined. This reveals a functional complementarity between the two groups: α-proteobacteria primarily participate in fundamental nitrogen and carbon cycling, while γ-proteobacteria dominate high-energy metabolic processes (Cheng et al., 2023; Luo et al., 2025). Oligotrophic conditions may promote the expansion of α-proteobacteria, whereas N- or P-rich environments might enhance the relative abundance of specific γ-proteobacteria lineages (Salamon Slater et al., 2023). Darlison et al. (2019), in their study on spinach (Spinacia oleracea) and wild rocket (Diplotaxis tenuifolia), confirmed that leaf mineral content influences phyllosphere microbial community assembly: with increasing N fertilizer application, bacterial community alpha diversity decreased, while changes in fungal communities varied by plant species. This study found that as leaf N and K concentrations decreased, bacterial α-diversity increased; conversely, fungal α-diversity were significantly affected by leaf P concentration.

4.2. Nutrient-mediated microbial assembly

Our finding indicates that interactions among phyllosphere microorganisms are often regulated by nutritional factors. Bacterial communities enhance the viability of other microbial groups by expanding cooperative networks, while nutrient-driven selection pressures foster more complex cooperative relationships, thereby improving system stability under nutrient-limited conditions (Fabiańska et al., 2019; Cheng and Ma, 2025). Conversely, nutrient enrichment can stimulate the rapid proliferation of specific bacterial taxa, leading to reduced microbial diversity, altered network structure, and ultimately, the destabilization of the network (Raimi et al., 2023). We found that bacterial network complexity varied with developmental stage and increased as leaf N/K contents decreased. In contrast, fungal network complexity was relatively low during the early and mid-stages characterized by higher nutrient availability but gradually increased, reaching a higher level of structural complexity in the final stage as nutrient levels declined. Our study showed that fungal network complexity was significantly correlated with leaf N/K contents. Nutrient factors play a crucial role in shaping fungal community structure, significantly affecting fungal community composition, abundance, diversity, and network complexity (Kerekes et al., 2013). These findings support the core argument that nutritional factors, particularly leaf N and P, drive structural changes in the phyllosphere fungal network, where reduced nutrient availability can promote increased fungal network complexity and diversity.

Microbial community assembly results from the combined influence of both deterministic and stochastic mechanisms. The relative significance of these processes is context-dependent, shifting across different habitats and under varying environmental conditions (Wang et al., 2023). Our study showed that deterministic mechanisms played the predominant role in shaping bacterial community assembly. Coincidentally, during the color-changing and stable stages on phylloplane when leaf nutrient factors were low, the bacterial communities were influenced by stochastic processes. This phenomenon suggests that the assembly mechanisms of microbial communities in the phyllosphere may also undergo significant changes in response to variations in environmental conditions (Luan et al., 2020). This aligns with the global-scale finding by Li et al. (2023), who showed that environmental filtering is the key determinant of plant-beneficial bacteria distribution. Through core taxa, bacteria formed stable networks with gradually increasing connectivity and complexity, while maintaining a certain degree of physiological isolation from the fluctuating below-ground niche. This strategy maximizes competitive efficiency and maintains system stability, enabling bacteria to rapidly utilize abundant environmental resources and gain a competitive advantage (He et al., 2025). In contrast, although fungal community assembly was also driven by deterministic processes, its network structure underwent significant fluctuations: the number of edges continuously increased during the early and middle stages but showed a declining trend in the final stage. The number of modules exhibited a “deep V-shaped” trajectory, stabilizing at a relatively high number after the color-changing stage, with negative correlations showing explosive growth during the light-green stage. By colonizing early, competitively expanding in the middle stage, and specializing in niches in the later stage, fungi more efficiently utilize below-ground nutrients for survival and reproduction in resource-scarce environments (Song et al., 2017), demonstrating stronger micro-niche differentiation capabilities and ecological persistence compared to bacteria.

4.3. Growth-stage shifts in microbial transfer between plant niches

This study revealed distinct niche differentiation in microbial sources. Epiphytic and endophytic leaf communities of both bacteria and fungi consistently served as the primary sources and sinks for one another, a pattern that remained stable across all developmental stages. This finding aligns with current understanding: although epiphytic microbes can originate from soil, air, and seeds (Xu et al., 2022), and their diversity and abundance are influenced by leaf microtopography such as vein density and stomatal traits (Yan et al., 2022), the direct exchange between epiphytic and endophytic compartments remains the most immediate pathway shaping phyllosphere communities.

Building on this core pattern, the relative contributions of secondary source pools exhibited dynamic shifts across developmental stages. During the bronze stage, phylloplane bacteria and fungi received higher contributions from leaf endosphere and the root surface, while leaf endosphere derived more from phylloplane and branch endosphere. This may be associated with increased biomass and diversity of root-associated microbiomes in biodiverse plant communities (Eisenhauer et al., 2017). By the light-green stage, contributions from branch surface while leaf endosphere to phylloplane communities increased, and leaf endosphere showed greater input from branch surface. This transition likely reflects adaptive adjustments of plants to changing environmental conditions during growth (Bao et al., 2022).

Furthermore, we found that below-ground niches contributed significantly more to fungal communities than to bacterial communities. This disparity may stem from fundamental differences in dispersal mechanisms and colonization strategies between the two kingdoms. Bacteria primarily rely on passive dispersal through water films (Furuno et al., 2010) and aerosols (Harb et al., 2021), resulting in relatively limited dispersal ranges; consequently, local sources such as root and branch surface contribute more prominently to bacterial communities. In contrast, fungi exhibit greater dispersal capacity and broader ecological niches than bacteria (Bahram et al., 2018). Through their extensive hyphal networks (Rog et al., 2020), fungi establish tighter connections between below-ground and above-ground niches, playing unique roles in facilitating resource exchange among plants, maintaining ecosystem stability (Germain and Lutz, 2021), and supporting biodiversity (Sandoz et al., 2020). This distinction may also partially explain the complementary roles of below-ground and above-ground compartments in plant health and ecosystem functioning (Martins et al., 2024): below-ground microorganisms support plant growth by enhancing nutrient uptake and disease resistance, while phyllosphere microorganisms regulate photosynthesis to convert solar energy into chemical energy, synthesizing organic compounds that provide energy and defense mechanisms for the plant (McGrane and Beattie, 2017; Nakagami et al., 2024). Rhizosphere microorganisms primarily interact with plants through root exudates, whereas phyllosphere microorganisms colonize based on epiphytic properties (Bulgarelli et al., 2012).

5. Conclusion

Our study elucidated that the assembly of phyllosphere microbial communities in rubber trees exhibits pronounced stage-specificity and is mediated by nutrient availability. Phyllosphere microbiota assembly in rubber trees exhibits stage-specificity mediated by leaf nutrients: bacterial diversity peak at budding via N/P/K mediation, while fungal diversity peaks at budding/stable stages with N/P-regulated. Stochastic processes dominate epiphytic bacteria during color-changing/stable stages, whereas deterministic N/K-driven processes prevail elsewhere and for endophytes. Bacterial network complexity increased progressively with N/K decreasing. Fungal network complexity diminishes post-early stages despite initial N/P promotion, with fungi mediating cross-niche exchanges. The results demonstrate clear differences in the responses of bacteria and fungi to leaf nutritional factors and in their assembly mechanisms within epiphytic and endophytic niches.

CRediT authorship contribution statement

Jiao Sun: Data curation, Formal analysis, Software, Visualization, Investigation, Writing - original draft, Writing - review & editing. Ya-Qing Wei: Resources, Conceptualization, Supervision, Project administration, Funding acquisition, Visualization, Writing - review & editing. Yan-Jun Du: Conceptualization, Supervision, Project administration, Writing – review & editing. Guo-Yu Lan: Resources, Conceptualization, Supervision, Project administration, Funding acquisition, Writing - review & editing.

Data availability statement

The raw reads were deposited into the NCBI Sequence Read Archive (SRA) database (Accession Number: PRJNA1312237). The plot data supporting the findings of this study are publicly available in the ScienceDB repository at https://doi.org/10.57760/sciencedb.37882.

Funding

This work was supported by the National Natural Science Foundation of China (32271603), the Central Public-interest Scientific Institution Basal Research Fund (1630022024013), the Chinese Academy of Tropical Agricultural Sciences for Science and Technology Innovation Team of National Tropical Agricultural Science Center (NO. CATASCXTD202504), the Earmarked Fund for Chinese Agricultural Research Systems (CARS-33-ZP3), and the Central Public-interest Scientific Institution Basal Research Fund (1630022022003).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

We would also like to thank the Tropical Forestry Ecology Research Group for their assistance in sampling work.

Footnotes

Peer review under the responsibility of Editorial Office of Plant Diversity.

Appendix A

Supplementary data to this article can be found online at https://doi.org/10.1016/j.pld.2026.05.013.

Contributor Information

Yanjun Du, Email: yanjundu1981@gmail.com.

Guoyu Lan, Email: langyrri@163.com.

Appendix A. Supplementary data

The following are the Supplementary data to this article.

Multimedia component 1
mmc1.docx (14.6MB, docx)

figs1.

figs1

Open in a new tab

References

  1. Abdelfattah A., Wisniewski M., Schena L., et al. Experimental evidence of microbial inheritance in plants and transmission routes from seed to phyllosphere and root. Environ. Microbiol. 2021;23:2199–2214. doi: 10.1111/1462-2920.15392. [DOI] [PubMed] [Google Scholar]
  2. Adams R.I., Miletto M., Taylor J.W., et al. Dispersal in microbes: Fungi in indoor air are dominated by outdoor air and show dispersal limitation at short distances. ISME J. 2013;7:1262–1273. doi: 10.1038/ismej.2013.28. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bao L., Sun B., Liu J., et al. Leaf-Associated epiphytic fungi of Ginkgo biloba, Pinus bungeana and Sabina chinensis exhibit delicate seasonal variations. J. Fungi. 2022;8 doi: 10.3390/jof8060631. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Barbour K.M., Martiny J.B.H. Investigating eco-evolutionary processes of microbial community assembly in the wild using a model leaf litter system. ISME J. 2024;18 doi: 10.1093/ismejo/wrae043. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bodenhausen N., Horton M.W., Bergelson J. Bacterial communities associated with the leaves and the roots of Arabidopsis thaliana. PLoS One. 2013;8 doi: 10.1371/journal.pone.0056329. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bolyen E., Rideout J.R., Dillon M.R., et al. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat. Biotechnol. 2019;37:852–857. doi: 10.1038/s41587-019-0209-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bulgarelli D., Rott M., Schlaeppi K., et al. Revealing structure and assembly cues for Arabidopsis root-inhabiting bacterial microbiota. Nature. 2012;488:91–95. doi: 10.1038/nature11336. [DOI] [PubMed] [Google Scholar]
  8. Bahram M., Hildebrand F., Forslund S.K., et al. Structure and function of the global topsoil microbiome. Nature. 2018;560:233–237. doi: 10.1038/s41586-018-0386-6. [DOI] [PubMed] [Google Scholar]
  9. Cheng K., Li X., Tong M., et al. Integrated metagenomic and metaproteomic analyses reveal bacterial micro-ecological mechanisms in coral bleaching. mSystems. 2023;8 doi: 10.1128/msystems.00505-23. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Cheng Y., Ma Y. Plant growth-promoting bacteria reshape rhizosphere microbial networks and biochemical properties to drive sustainable Zea mays growth. Appl. Soil Ecol. 2025;213 doi: 10.1016/j.apsoil.2025.106275. [DOI] [Google Scholar]
  11. Chu X.-L., Zhang Q.-G., Buckling A., et al. Interspecific niche competition increases morphological diversity in multi-species microbial communities. Front. Microbiol. 2021;12 doi: 10.3389/fmicb.2021.699190. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Carnier D.M., Martinelli A.P. The rubber tree (Hevea brasiliensis Muell. Arg.) homologue of the leafy/floricaula gene is preferentially expressed in both male and female floral meristems. J. Exp. Bot. 2005;56:1965–1974. doi: 10.1093/jxb/eri194. [DOI] [PubMed] [Google Scholar]
  13. Darlison J., Mogren L., Rosberg A.K., et al. Leaf mineral content govern microbial community structure in the phyllosphere of spinach (Spinacia oleracea) and rocket (Diplotaxis tenuifolia) Sci. Total Environ. 2019;675:501–512. doi: 10.1016/j.scitotenv.2019.04.254. [DOI] [PubMed] [Google Scholar]
  14. Edgar R.C. Search and clustering orders of magnitude faster than BLAST. Bioinformatics. 2010;26:2460–2461. doi: 10.1093/bioinformatics/btq461. [DOI] [PubMed] [Google Scholar]
  15. Eisenhauer N., Lanoue A., Strecker T., et al. Root biomass and exudates link plant diversity with soil bacterial and fungal biomass. Sci. Rep. 2017;7 doi: 10.1038/srep44641. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Evans R., Beckerman A.P., Wright R.C.T., et al. Eco-evolutionary dynamics set the tempo and trajectory of metabolic evolution in multispecies communities. Curr. Biol. 2020;30:4984–4988.e4. doi: 10.1016/j.cub.2020.09.028. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Fabiańska I., Sosa-Lopez E., Bucher M. The role of nutrient balance in shaping plant root-fungal interactions: facts and speculation. Curr. Opin. Microbiol. 2019;49:90–96. doi: 10.1016/j.mib.2019.10.004. [DOI] [PubMed] [Google Scholar]
  18. Furuno S., Päzolt K., Rabe C., et al. Fungal mycelia allow chemotactic dispersal of polycyclic aromatic hydrocarbon-degrading bacteria in water-unsaturated systems. Environ. Microbiol. 2010;12:1391–1398. doi: 10.1111/j.1462-2920.2009.02022.x. [DOI] [PubMed] [Google Scholar]
  19. Gao J., Uwiringiyimana E., Zhang D. Microbial composition and diversity of the tobacco leaf phyllosphere during plant development. Front. Microbiol. 2023;14 doi: 10.3389/fmicb.2023.1199241. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Germain S.J., Lutz J.A. Shared friends counterbalance shared enemies in old forests. Ecology. 2021;102 doi: 10.1002/ecy.3495. [DOI] [PubMed] [Google Scholar]
  21. Hao Y., Zhao X., Zhang D. Field experimental evidence that stochastic processes predominate in the initial assembly of bacterial communities. Environ. Microbiol. 2016;18:1730–1739. doi: 10.1111/1462-2920.12858. [DOI] [PubMed] [Google Scholar]
  22. He D., Dai Z., Cheng S., et al. Microbial life-history strategies and genomic traits between pristine and cropland soils. mSystems. 2025;10 doi: 10.1128/msystems.00178-25. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Huang H. 2021. Linket: Everything is Linkable.https://github.com/Hy4m/linkET [Google Scholar]
  24. Harb C., Pan J., DeVilbiss S., et al. Increasing freshwater salinity impacts aerosolized bacteria. Environ. Sci. Technol. 2021;55:5731–5741. doi: 10.1021/acs.est.0c08558. [DOI] [PubMed] [Google Scholar]
  25. Jeraldo P., Sipos M., Chia N., et al. Quantification of the relative roles of niche and neutral processes in structuring gastrointestinal microbiomes. Proc. Natl. Acad. Sci. U.S.A. 2012;109:9692–9698. doi: 10.1073/pnas.1206721109. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Jiao S., Lu Y., Wei G. Soil multitrophic network complexity enhances the link between biodiversity and multifunctionality in agricultural systems. Glob. Change Biol. 2022;28:140–153. doi: 10.1111/gcb.15917. [DOI] [PubMed] [Google Scholar]
  27. Kalyuzhny M., Seri E., Chocron R., et al. Niche versus neutrality: a dynamical analysis. Am. Nat. 2014;184:439–446. doi: 10.1086/677930. [DOI] [PubMed] [Google Scholar]
  28. Ke-chang N., Yi-ning L., Ze-hao S., et al. Community assembly: the relative importance of neutral theory and niche theory. Biodivers. Sci. 2009;17 doi: 10.3724/SP.J.1003.2009.09142. [DOI] [Google Scholar]
  29. Kerekes J., Kaspari M., Stevenson B., et al. Nutrient enrichment increased species richness of leaf litter fungal assemblages in a tropical forest. Mol. Ecol. 2013;22:2827–2838. doi: 10.1111/mec.12259. [DOI] [PubMed] [Google Scholar]
  30. Khandoori R., Mondal K., Ghosh P. Resource limitation and population fluctuation drive spatiotemporal order in microbial communities. Soft Matter. 2024;20:3823–3835. doi: 10.1039/D4SM00066H. [DOI] [PubMed] [Google Scholar]
  31. Kimura M. Recent development of the neutral theory viewed from the Wrightian tradition of theoretical population genetics. Proc. Natl. Acad. Sci. U.S.A. 1991;88:5969–5973. doi: 10.1073/pnas.88.14.5969. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Laforest-Lapointe I., Messier C., Kembel S.W. Tree leaf bacterial community structure and diversity differ along a gradient of urban intensity. mSystems. 2017;2 doi: 10.1128/mSystems.00087-17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Lan G., Wei Y., Li Y., et al. Diversity and assembly of root-associated microbiomes of rubber trees. Front. Plant Sci. 2023;14 doi: 10.3389/fpls.2023.1136418. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Lan G., Wei Y., Zhang X., et al. Assembly and maintenance of phyllosphere microbial diversity during rubber tree leaf senescence. Commun. Biol. 2024;7 doi: 10.1038/s42003-024-06907-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Li J., Li C., Kou Y., et al. Distinct mechanisms shape soil bacterial and fungal co-occurrence networks in a mountain ecosystem. FEMS Microbiol. Ecol. 2020;96 doi: 10.1093/femsec/fiaa030. [DOI] [PubMed] [Google Scholar]
  36. Li P., Tedersoo L., Crowther T.W., et al. Fossil-fuel-dependent scenarios could lead to a significant decline of global plant-beneficial bacteria abundance in soils by 2100. Nat. Food. 2023;4:996–1006. doi: 10.1038/s43016-023-00869-9. [DOI] [PubMed] [Google Scholar]
  37. Li L., Zheng R., Wang Z., et al. Leaf health status regulates endophytic microbial community structure, network complexity, and assembly processes in the leaves of the rare and endangered plant species Abies fanjingshanensis. Microorganisms. 2024;12 doi: 10.3390/microorganisms12071254. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Li Y., Lan G., Xia Y. Rubber trees demonstrate a clear retranslocation under seasonal drought and cold stresses. Front. Plant Sci. 2016;7 doi: 10.3389/fpls.2016.01907. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Lindow S.E., Brandl M.T. Microbiology of the phyllosphere. Appl. Environ. Microbiol. 2003;69:1875–1883. doi: 10.1128/AEM.69.4.1875-1883.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Lindow S.E., Leveau J.H.J. Phyllosphere microbiology. Curr. Opin. Biotechnol. 2002;13:238–243. doi: 10.1016/S0958-1669(02)00313-0. [DOI] [PubMed] [Google Scholar]
  41. Luan L., Liang C., Chen L., et al. Coupling bacterial community assembly to microbial metabolism across soil profiles. mSystems. 2020;5 doi: 10.1128/msystems.00298-20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Luo C., Zhang T., Mustafa M.F., et al. Removal efficiency of ARGs in different wastewater treatment plants and their potential risks in effluent. npj Clean Water. 2025;8 doi: 10.1038/s41545-025-00456-4. [DOI] [Google Scholar]
  43. Martins C.S.C., Delgado-Baquerizo M., Jayaramaiah R.H., et al. Aboveground and belowground biodiversity have complementary effects on ecosystem functions across global grasslands. PLoS Biology. 2024;22 doi: 10.1371/journal.pbio.3002736. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Masclaux C., Valadier M.-H., Brugière N., et al. Characterization of the sink/source transition in tobacco (Nicotiana tabacum L.) shoots in relation to nitrogen management and leaf senescence. Planta. 2000;211:510–518. doi: 10.1007/s004250000310. [DOI] [PubMed] [Google Scholar]
  45. McGrane R., Beattie G.A. Pseudomonas syringae pv. Syringae B728a regulates multiple stages of plant colonization via the bacteriophytochrome BphP1. mBio. 2017;8 doi: 10.1128/mBio.01178-17. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Mendes R., Kruijt M., De Bruijn I., et al. Deciphering the rhizosphere microbiome for disease-suppressive bacteria. Science. 2011;332:1097–1100. doi: 10.1126/science.1203980. [DOI] [PubMed] [Google Scholar]
  47. Millard P., Grelet G.-a. Nitrogen storage and remobilization by trees: ecophysiological relevance in a changing world. Tree Physiol. 2010;30:1083–1095. doi: 10.1093/treephys/tpq042. [DOI] [PubMed] [Google Scholar]
  48. Muneer M.A., Huang R., Xiaojun Y., et al. Mg-enriched nutrient management enhances phyllosphere bacterial diversity, community structure, and functional traits in pomelo orchards. Curr. Res. Microb. Sci. 2025;9 doi: 10.1016/j.crmicr.2025.100476. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Nakagami S., Wang Z., Han X., et al. Regulation of bacterial growth and behavior by host plant. Annu. Rev. Phytopathol. 2024;62:69–96. doi: 10.1146/annurev-phyto-010824-023359. [DOI] [PubMed] [Google Scholar]
  50. Ning D., Deng Y., Tiedje J.M., et al. A general framework for quantitatively assessing ecological stochasticity. Proc. Natl. Acad. Sci. U.S.A. 2019;116:16892–16898. doi: 10.1073/pnas.1904623116. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Oksanen J., Blanchet G.F., Kindt R., et al. 2013. Vegan: Community Ecology Package. R Package Version 2.0-10. the R Project. [Google Scholar]
  52. Peipoch M., Miller S.R., Antao T.R., et al. Niche partitioning of microbial communities in riverine floodplains. Sci. Rep. 2019;9 doi: 10.1038/s41598-019-52865-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Peng C. 2025. MetaNet: Network Analysis for Omics Data.https://CRAN.R-project.org/package=MetaNet [Google Scholar]
  54. Purahong W., Wubet T., Lentendu G., et al. Life in leaf litter: novel insights into community dynamics of bacteria and fungi during litter decomposition. Mol. Ecol. 2016;25:4059–4074. doi: 10.1111/mec.13739. [DOI] [PubMed] [Google Scholar]
  55. R Core Team . R Foundation for Statistical Computing; 2024. R: a Language and Environment for Statistical Computing.https://www.R-project.org/ [Google Scholar]
  56. Raimi A.R., Ezeokoli O.T., Adeleke R.A. Soil nutrient management influences diversity, community association and functional structure of rhizosphere bacteriome under vegetable crop production. Front. Microbiol. 2023;14 doi: 10.3389/fmicb.2023.1229873. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Rog I., Rosenstock N.P., Körner C., et al. Share the wealth: trees with greater ectomycorrhizal species overlap share more carbon. Mol. Ecol. 2020;29:2321–2333. doi: 10.1111/mec.15351. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Ruiz-Pérez C.A., Restrepo S., Zambrano M.M. Microbial and functional diversity within the phyllosphere of Espeletia species in an Andean high-mountain ecosystem. Appl. Environ. Microbiol. 2016;82:1807–1817. doi: 10.1128/AEM.02781-15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Salamon Slater E.R., Turk-Kubo K.A., Hallstrøm S., et al. Composition and distribution of diazotrophs in the Baltic Sea. Estuar. Coast Shelf Sci. 2023;294 doi: 10.1016/j.ecss.2023.108527. [DOI] [Google Scholar]
  60. Sampson T.R., Debelius J.W., Thron T., et al. Gut microbiota regulate motor deficits and neuroinflammation in a model of Parkinson’s disease. Cell. 2016;167:1469–1480.e12. doi: 10.1016/j.cell.2016.11.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Sandoz F.A., Bindschedler S., Dauphin B., et al. Biotic and abiotic factors shape arbuscular mycorrhizal fungal communities associated with the roots of the widespread fern Botrychium lunaria (Ophioglossaceae) Environ. Microbiol. Rep. 2020;12:342–354. doi: 10.1111/1758-2229.12840. [DOI] [PubMed] [Google Scholar]
  62. Song H.-K., Song W., Kim M., et al. Bacterial strategies along nutrient and time gradients, revealed by metagenomic analysis of laboratory microcosms. FEMS Microbiol. Ecol. 2017;93 doi: 10.1093/femsec/fix114. fix114. [DOI] [PubMed] [Google Scholar]
  63. Soria-Píriz S., Corzo A., Jiménez-Arias J.L., et al. Niche selection in bacterioplankton: a study of taxonomic composition and single-cell characteristics in an acidic reservoir. Environ. Microbiol. Rep. 2024;16 doi: 10.1111/1758-2229.13255. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Su K., Yu D., Wu H. Accurate feature selection for single-cell clustering. FEAST. 2021 doi: 10.1093/bib/bbab034. https://bioc.r-universe.dev/FEAST [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Vorholt J.A. Microbial life in the phyllosphere. Nat. Rev. Microbiol. 2012;10:828–840. doi: 10.1038/nrmicro2910. [DOI] [PubMed] [Google Scholar]
  66. Voříšková J., Baldrian P. Fungal community on decomposing leaf litter undergoes rapid successional changes. ISME J. 2013;7:477–486. doi: 10.1038/ismej.2012.116. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Vujanovic V. Tremellomycetes yeasts in kernel ecological niche: early indicators of enhanced competitiveness of endophytic and mycoparasitic symbionts against wheat Pathobiota. Plants. 2021;10 doi: 10.3390/plants10050905. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Walters W., Hyde E.R., Berg-Lyons D., et al. Improved bacterial 16S rRNA gene (V4 and V4-5) and Fungal internal transcribed spacer marker gene primers for microbial community surveys. mSystems. 2016;1 doi: 10.1128/mSystems.00009-15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Wang J., Pan Z., Yu J., et al. Global assembly of microbial communities. mSystems. 2023;8 doi: 10.1128/msystems.01289-22. [DOI] [PMC free article] [PubMed] [Google Scholar]
  70. Wang L., Liu Z., Bres C., et al. Coniferous tree species identity and leaf aging alter the composition of Phyllosphere communities through changes in leaf traits. Microb. Ecol. 2024;87 doi: 10.1007/s00248-024-02440-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Weatherhead E., Davis E.L., Koide R.T. Many foliar endophytic fungi of Quercus gambelii are capable of psychrotolerant saprotrophic growth. PLoS One. 2022;17 doi: 10.1371/journal.pone.0275845. [DOI] [PMC free article] [PubMed] [Google Scholar]
  72. Wei Y., Lan G., Wu Z., et al. Phyllosphere fungal communities of rubber trees exhibited biogeographical patterns, but not bacteria. Environ. Microbiol. 2022;24:3777–3790. doi: 10.1111/1462-2920.15894. [DOI] [PubMed] [Google Scholar]
  73. Wei Y.Q., Zhang X.C., Peng W.T., et al. Temperature and rainfall mediate soil-plant microbiome assembly in tropical rubber plantations. Appl. Soil Ecol. 2025;216 doi: 10.1016/j.apsoil.2025.106532. [DOI] [Google Scholar]
  74. Wickham H. second ed. Springer; 2016. ggplot2: Elegant Graphics for Data Analysis. [Google Scholar]
  75. Xiong C., He J., Singh B.K., et al. Rare taxa maintain the stability of crop mycobiomes and ecosystem functions. Environ. Microbiol. 2021;23:1907–1924. doi: 10.1111/1462-2920.15262. [DOI] [PubMed] [Google Scholar]
  76. Xiong C., Singh B.K., He J.-Z., et al. Plant developmental stage drives the differentiation in ecological role of the maize microbiome. Microbiome. 2021;9 doi: 10.1186/s40168-021-01118-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  77. Xu N., Zhao Q., Zhang Z., et al. Phyllosphere microorganisms: sources, drivers, and their interactions with plant hosts. J. Agric. Food Chem. 2022;70:11081–11098. doi: 10.1021/acs.jafc.2c04233. [DOI] [PubMed] [Google Scholar]
  78. Xu W., Ci X., Song C., et al. Soil phosphorus heterogeneity promotes tree species diversity and phylogenetic clustering in a tropical seasonal rainforest. Ecol. Evol. 2016;6:8719–8726. doi: 10.1002/ece3.2529. [DOI] [PMC free article] [PubMed] [Google Scholar]
  79. Yan K., Han W., Zhu Q., et al. Leaf surface microtopography shaping the bacterial community in the phyllosphere: evidence from 11 tree species. Microbiol. Res. 2022;254 doi: 10.1016/j.micres.2021.126897. [DOI] [PubMed] [Google Scholar]
  80. Yang M., Wang J., Qi Y., et al. Plant developmental stage drives the assembly and functional adaptability of endophytic microbial communities. Front. Microbiol. 2025;16 doi: 10.3389/fmicb.2025.1492141. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Multimedia component 1
mmc1.docx (14.6MB, docx)

Data Availability Statement

The raw reads were deposited into the NCBI Sequence Read Archive (SRA) database (Accession Number: PRJNA1312237). The plot data supporting the findings of this study are publicly available in the ScienceDB repository at https://doi.org/10.57760/sciencedb.37882.

Articles from Plant Diversity are provided here courtesy of Kunming Institute of Botany, Chinese Academy of Sciences

RESOURCES