Effects of Variety, Plant Location, and Season on the Phyllosphere Bacterial Community Structure of Alfalfa (Medicago sativa L.)

Plant phyllosphere bacteria are vital for plant health and productivity and are affected by both abiotic and biotic factors. In this study, we surveyed the structure of the phyllosphere bacterial community associated with alfalfa. For two varieties of alfalfa, forty-eight samples of phyllosphere communities were collected at two locations over four seasons in 2020. Proteobacteria and actinobacteria were associated with the dominating phylum in the bacterial communities of the alfalfa phyllosphere. Sphingomonas was the most abundant genus-level bacteria, followed by Methylobacterium, Burkholderia-Caballeronia-Paraburkholderia, and Pseudomonas. Sampling time had a greater affect than site and variety on alfalfa surface microorganisms. The variation in phyllosphere bacterial community assembly was mostly explained by the season–site interaction (43%), season–variety interaction (35%), and season (28%). Variety, site–variety interaction, and season–site–variety interactions did not have a meaningful effect on phyllosphere bacterial diversity and community structure. The bacterial community in the phyllosphere of alfalfa showed seasonal changes over time. The environmental factors that contributed most to the phyllosphere bacterial community of alfalfa were temperature and sunshine duration, which were significantly positively correlated with most of the dominant bacterial genera in the alfalfa phyllosphere.


Background
The entire outer surface of the aboveground portion of a plant is referred to as the phyllosphere. This surface offers a habitat favorable for the survival of some microorganisms [1]. There are rich and diverse microbial communities in the phyllospheres of forage plants, mainly dominated by bacterial communities [2]. On average, the phyllosphere can support 10 6~1 0 7 bacterial cells per square centimeter, with bacteria occupying a major position in the phyllosphere microbial community [3]. These phyllosphere microorganisms are involved in the physiological metabolism of forage grasses through interactions with the forage host [4]. They play an important role in the global carbon and nitrogen cycles and have not only a complex relationship with plant hosts but also various functions, such as fixing nitrogen [5], promoting forage growth [6], enhancing forage resistance [7], changing forage surface properties, and degrading pesticide residues on forage surfaces [8,9].
The phyllosphere environment is a very demanding and unstable habitat for phyllosphere microorganisms, which live on the plant surface and are often exposed to water and nutrient deficiencies as well as drastic environmental changes, including bright light, high temperatures [10], and high UV exposure [11]. In addition, biological factors such as plant species and plant physiological status also affect phyllosphere microorganisms [12]. Different plants can selectively accumulate various microorganisms in the rhizosphere, leaf space, and inner layers. Due to the significant role that plant-associated microbiota play in the regulation of numerous biological processes, as well as their influence on a

Study Site
The study plots were located in two cities in Guizhou Province, China: Guiyang, at the Guizhou University pilot site (106 • 07 E; 26 • 11 N; elevation 1100 m), and Tongren (108 • 20 14 E; 27 • 32 03 N; elevation 614 m). The distance between sites is 248 km. Both experimental sites had not been planted with alfalfa, the Guiyang experimental site had not been planted with crops in the early stage, mainly wild weeds, and the Tongren experimental site had corn as the previous crop. The data of soil basic nutrients in Guiyang and Tongren experimental sites were provided by the research group of Mr. Dong of Guizhou University (Appendix A) [33]. Two alfalfa varieties, Medicago sativa L. cv. Xinjiang Daye, and Medicago sativa L. cv. Algonguin, which are suitable for growing in acidic soils in the humid and hot regions of southwest China, were selected for planting in each trial field. Each site had 3 plots, and 60 individual plants of each variety were planted in each plot, with 50 cm intervals between plant rows. Alfalfa was planted at the start of 2018, mown six times a year from 2019 onward, and sampled at the initial flowering stage prior to mowing. No fertilization and artificial irrigation treatments were conducted. The experimental field of Guizhou University in Guiyang has a subtropical humid monsoon climate, with obvious plateau climate characteristics. The experimental field in Tongren belongs to the central subtropical humid monsoon climate, with sufficient sunshine and abundant rainfall. We obtained monthly climate data from the China National Meteorological Information Center ( Figure 1).
The study plots were located in two cities in Guizhou Province, China: Guiyang, at the Guizhou University pilot site (106°07′ E; 26°11′ N; elevation 1100 m), and Tongren (108°20′14″ E; 27°32′03″ N; elevation 614 m). The distance between sites is 248 km. Both experimental sites had not been planted with alfalfa, the Guiyang experimental site had not been planted with crops in the early stage, mainly wild weeds, and the Tongren experimental site had corn as the previous crop. The data of soil basic nutrients in Guiyang and Tongren experimental sites were provided by the research group of Mr. Dong of Guizhou University (Appendix A) [33]. Two alfalfa varieties, Medicago sativa L. cv. Xinjiang Daye, and Medicago sativa L. cv. Algonguin, which are suitable for growing in acidic soils in the humid and hot regions of southwest China, were selected for planting in each trial field. Each site had 3 plots, and 60 individual plants of each variety were planted in each plot, with 50 cm intervals between plant rows. Alfalfa was planted at the start of 2018, mown six times a year from 2019 onward, and sampled at the initial flowering stage prior to mowing. No fertilization and artificial irrigation treatments were conducted. The experimental field of Guizhou University in Guiyang has a subtropical humid monsoon climate, with obvious plateau climate characteristics. The experimental field in Tongren belongs to the central subtropical humid monsoon climate, with sufficient sunshine and abundant rainfall. We obtained monthly climate data from the China National Meteorological Information Center (Figure 1).

Bacterial Community Collection
A total of 48 samples were collected. We sampled phyllosphere bacteria from these two varieties of alfalfa (Xinjiang Daye and Algonquin) at two sites (Guiyang and Tongren) during four seasons (January, April, July, and October) in 2020. Each site contained three planting plots, five sampling points were randomly selected within each plot (each 2 m apart), and three alfalfa plants were randomly selected for sampling at each sampling point. Each alfalfa plant was swabbed with 5 swabs, and 75 swabs per plot were collected together as one sample for bacterial sequencing. The cotton swab sampling method was

Bacterial Community Collection
A total of 48 samples were collected. We sampled phyllosphere bacteria from these two varieties of alfalfa (Xinjiang Daye and Algonquin) at two sites (Guiyang and Tongren) during four seasons (January, April, July, and October) in 2020. Each site contained three planting plots, five sampling points were randomly selected within each plot (each 2 m apart), and three alfalfa plants were randomly selected for sampling at each sampling point. Each alfalfa plant was swabbed with 5 swabs, and 75 swabs per plot were collected together as one sample for bacterial sequencing. The cotton swab sampling method was used for bacterial community collection [34]. Sterile cotton swabs (Wela Biotechnology Co., Ltd., Guiyang, China) were moistened with sterilized saline (0.9% concentration) and wiped on the outer surface of the aboveground portion of the alfalfa (stems, leaves) for 10 s. The heads of the swabs were collected and quickly placed in sterilized centrifuge tubes, snap-frozen in liquid nitrogen, and brought back to the laboratory for storage at −80 • C.

Microbiome Analysis
Using a HiPure Soil DNA Extraction Kit, microbial DNA was isolated (Magen, Guangzhou, China). Using the following parameters, the 16S rRNA gene V3-V4 region was amplified with PCR: 94 • C for 2 min, 98 • C for 10 s, 62 • C for 30 s, and 68 • C for 30 s for  30 cycles, and lastly 68 • C for 5 min. This process was accomplished using the primers 341F: CCTACGGGNGGCWGCAG and 806 R: GGACTACHVGGGTATCTAAT [35]. PCRs were carried out three times. The following was the composition of the amplification system: 50 L of a mixture of 5 L of KOD buffer (10 M), 5 L of 2 mm dNTPs, 3 L of 25 mm MgSO4, 1.5 L of forward and reverse primers (10 M), 1 L of KOD polymerase (TOYOBO, Osaka, Japan), and 100 ng of template DNA.
Using the AxyPrep DNA Gel Extraction Kit (Axygen Biosciences, Union City, CA, USA), amplicons were extracted from 2% agarose gels, purified in accordance with the manufacturer's instructions, and quantified using the ABI StepOnePlus Real-Time PCR System (Life Technologies, Foster City, CA, USA). On an Illumina platform, the purified amplicons were sequenced in accordance with best practices, then paired-end sequencing (PE250) was carried out in accordance with best practices.
The clean readings, which had a minimum overlap of 10 base pairs and a maximum mismatch rate of 2%, were combined into tags using FLASH [36] (version 1.2.11) via the following procedure: when the number of bases in the continuous low-quality value (the default quality threshold) reached the predetermined length (the default length is 3 bp), raw tags from the first low-quality base site were broken, and tags with a continuous high-quality base length less than 75% of the tag length were filtered. Using the UCHIME technique, tags' chimeras (for 16S sequencing analysis) were examined [37]. For further study, the clean tags that were acquired after filtering the chimeras were employed. Using UPARSE [38] (version 9.2.64), the clean tags were organized into operational taxonomic units (OTUs) based on a 97% similarity threshold. As the representative sequences for each OTU, the tag sequences with the highest abundance were chosen. Representative OTU sequences and the SILVA [39] database were compared (version 132).

Statistical Analysis
Species abundance chorograms, i.e., presented in a circular layout, were plotted using circus software [40] (version 0.69-3). The vegan package in R (version 2.5.3) was used to perform beta diversity analysis for NMDS multivariate statistical analysis based on Bray-Curtis distances. For Bray-Curtis distances, the PERMANOVA test was used. R (version 2.5.3)'s vegan package was used to perform redundancy analysis (RDA) to elucidate how environmental factors will affect community composition in 2020. Using the R language's psych package, the Pearson's correlation coefficients between environmental parameters and species were determined (version 1.8.4). Utilizing the free online platform of Omicshare tools, the data were evaluated.

Phyllosphere Bacterial Diversity and Bacterial Community Dynamics
A total of 4,997,132 high-quality sequences were produced after quality filtering and the elimination of nonbacterial sequence reads. As seen in Figure 2, as the quantity of sequencing data reached approximately 20,000, the rarefaction curve of each sample tended to approach the plateau phase. It was clear from this that there were enough sequences in these samples to adequately describe the bacterial composition.
In the alfalfa samples, sequencing revealed 7393 bacterial operational taxonomic units (OTUs, binned at 97% similarity). The sampling depth effectively captured the majority of the bacterial communities, as shown by the Good's coverage of all samples, which was roughly 0.99. The OTUs, Shannon index, Chao index, and Ace index tended to increase from spring to autumn and then decrease from autumn to winter in all alfalfa samples, except for the Xinjiang Daye group grown in Guiyang, but these trends were not significant (Table 1). tended to approach the plateau phase. It was clear from this that there were enough sequences in these samples to adequately describe the bacterial composition. In the alfalfa samples, sequencing revealed 7393 bacterial operational taxonomic units (OTUs, binned at 97% similarity). The sampling depth effectively captured the majority of the bacterial communities, as shown by the Good's coverage of all samples, which was roughly 0.99. The OTUs, Shannon index, Chao index, and Ace index tended to increase from spring to autumn and then decrease from autumn to winter in all alfalfa samples, except for the Xinjiang Daye group grown in Guiyang, but these trends were not significant (Table 1). The bacterial community structures of all alfalfa samples were illustrated using an NMDS plot based on the Bray-Curtis distance ( Figure 3). The samples in the seasons were obviously distinguished, indicating that the alfalfa phyllosphere had similar bacte-  The bacterial community structures of all alfalfa samples were illustrated using an NMDS plot based on the Bray-Curtis distance ( Figure 3). The samples in the seasons were obviously distinguished, indicating that the alfalfa phyllosphere had similar bacterial communities at each season regardless of the variety and plant location. However, Figure 3 shows that the different varieties of alfalfa in Guiyang and Tongren exhibited overlap to a certain extent in spring and summer and a certain extent in autumn and winter.
rial communities at each season regardless of the variety and plant location. However, Figure 3 shows that the different varieties of alfalfa in Guiyang and Tongren exhibited overlap to a certain extent in spring and summer and a certain extent in autumn and winter. The relative abundance of the alfalfa phyllosphere bacterial microbiome is shown in Figure 4. The 16S ribosomal DNA analysis demonstrated that, according to the composition of the phyllosphere community, Proteobacteria and Actinobacteria were the most dominant bacterial phyla. From spring to winter, Proteobacteria showed an increasing and then decreasing trend in all taxa and reached maximum relative abundance during the summer. The relative abundance of Actinobacteria showed a decreasing and then increasing trend for Algonguin in the four seasons at the two sites; however, a continuous decreasing trend was observed in Xinjiang Daye. In addition, many other phyla, such as Bacteroidetes (1.10% to 19.22%), Firmicutes (2.50% to 47.29%), and Planctomycetes (0.31% to 5.35%), were identified. Their relative abundance in all samples showed variable changes.
The relative abundance of the top 10 bacterial genera in the samples is shown in Figure 4B. Sphingomonas, Methylobacterium, Burkholderia-Caballeronia-Paraburkholderia, and Pseudomonas were the top genera. Among them, Sphingomonas, as a dominant genus, appeared more abundant in summer, and Methylobacterium showed higher relative abundance in winter in Guiyang and in autumn in Tongren. The relative abundance of the alfalfa phyllosphere bacterial microbiome is shown in Figure 4. The 16S ribosomal DNA analysis demonstrated that, according to the composition of the phyllosphere community, Proteobacteria and Actinobacteria were the most dominant bacterial phyla. From spring to winter, Proteobacteria showed an increasing and then decreasing trend in all taxa and reached maximum relative abundance during the summer. The relative abundance of Actinobacteria showed a decreasing and then increasing trend for Algonguin in the four seasons at the two sites; however, a continuous decreasing trend was observed in Xinjiang Daye. In addition, many other phyla, such as Bacteroidetes (1.10% to 19.22%), Firmicutes (2.50% to 47.29%), and Planctomycetes (0.31% to 5.35%), were identified. Their relative abundance in all samples showed variable changes.
The relative abundance of the top 10 bacterial genera in the samples is shown in Figure 4B. Sphingomonas, Methylobacterium, Burkholderia-Caballeronia-Paraburkholderia, and Pseudomonas were the top genera. Among them, Sphingomonas, as a dominant genus, appeared more abundant in summer, and Methylobacterium showed higher relative abundance in winter in Guiyang and in autumn in Tongren.

Drivers of Variation in Phyllosphere Bacterial Community Composition and Diversity
According to a PERMANOVA of community structure variation, the interaction between season and site explained 42.9% of the variation in bacterial communities, the interaction between season and variety explained 34.7%, season explained 27.6%, and site explained 4.0%. However, neither variety, nor the interaction of variety and site, nor the interaction of season, site, and variety had a significant influence on the diversity and community structure of phyllosphere bacteria (Table 2).

Drivers of Variation in Phyllosphere Bacterial Community Composition and Diversity
According to a PERMANOVA of community structure variation, the interaction between season and site explained 42.9% of the variation in bacterial communities, the interaction between season and variety explained 34.7%, season explained 27.6%, and site explained 4.0%. However, neither variety, nor the interaction of variety and site, nor the interaction of season, site, and variety had a significant influence on the diversity and community structure of phyllosphere bacteria (Table 2). Examining the relationship between microorganisms and environmental factors is useful for comprehending the mechanism of phyllosphere community assembly of alfalfa. Distance-based redundancy analysis (db-RDA) and a correlation analysis between microbiota and driving factors were carried out to clarify the primary environmental drivers in various alfalfa phyllospheres under various groups, as shown in Figure 5A. The results showed that the explanatory rate of environmental factors for the distribution of bacterial communities was 83.08%. This indicated that environmental factors had an important impact on the phyllosphere microorganisms of alfalfa. In addition, temperature, sunshine duration, precipitation, and relative humidity had a strong correlation with the microbial community (p < 0.05). Precipitation, temperature, and sunshine duration were positively correlated with most alfalfa bacterial communities sampled in summer. The relative humidity was strongly correlated with the phyllosphere bacterial community of alfalfa sampled in autumn and winter.
important impact on the phyllosphere microorganisms of alfalfa. In addition, tempera ture, sunshine duration, precipitation, and relative humidity had a strong correlatio with the microbial community (p < 0.05). Precipitation, temperature, and sunshine dura tion were positively correlated with most alfalfa bacterial communities sampled i summer. The relative humidity was strongly correlated with the phyllosphere bacteria community of alfalfa sampled in autumn and winter.
We used Pearson coefficients to investigate the correlations between specific gener and driving factors to better understand these linkages (p < 0.05; Figure 5C). Figure 5C shows that the genera Sphingomonas, Pseudomonas, and Pantoea were positively correlate with temperature, precipitation, and sunshine duration. This explains the higher relativ abundance of these three bacterial genera in summer than in other seasons. We als found that temperature, precipitation, and sunshine duration were negatively correlated with Bacillus, and relative humidity was negatively correlated with Sphingomonas, Pseu domonas, Pantoea, and Bacillus. The environmental factors that contributed most to th phyllosphere bacterial community of alfalfa were temperature and sunshine duratio ( Figure 5B).  We used Pearson coefficients to investigate the correlations between specific genera and driving factors to better understand these linkages (p < 0.05; Figure 5C). Figure 5C shows that the genera Sphingomonas, Pseudomonas, and Pantoea were positively correlated with temperature, precipitation, and sunshine duration. This explains the higher relative abundance of these three bacterial genera in summer than in other seasons. We also found that temperature, precipitation, and sunshine duration were negatively correlated with Bacillus, and relative humidity was negatively correlated with Sphingomonas, Pseudomonas, Pantoea, and Bacillus. The environmental factors that contributed most to the phyllosphere bacterial community of alfalfa were temperature and sunshine duration ( Figure 5B).

Analysis of Bacterial Community Composition and Diversity
According to DNA sequencing, Proteobacteria were the most prevalent phylum, followed by Actinobacteria. The most prevalent phyllosphere bacteria in a variety of crops, including miscanthus [41], thale cress [42], tea [43], and rice [44], are Proteobacteria and Actinobacteria. Proteobacteria are able to colonize a variety of habitats, including the rhizosphere and phyllosphere, which could account for their predominate presence [45].
The relative abundance of dominant genera of alfalfa phyllosphere bacteria in different locations, varieties, and sampling times is not very high. This might be because the phyllosphere microbial community is also characterized by a low number of highly abundant microbial taxa and a high number of rare taxa with low relative abundance [3,46]. The major bacteria at the genus level were Sphingomonas, Methylobacterium, Burkholderia-Caballeronia-Paraburkholderia, and Pseudomonas. All these genera were found to be dominant in different plant phyllospheres [19,47,48]. Notably, differences still existed, while certain dominant bacterial genera appear to be the same in other studies on alfalfa phyllosphere bacteria. Mc Garvey et al. [49] found that the main alfalfa phyllosphere genera were Erwinia, Escherichia, Pseudomonas, Pantoea, and Enterobacter, while Zheng et al. [50] found the main bacterial genera to be Pantoea, Enterobacter, and Buchnera. In combination with our experimental results, the same suggestion was made that different seasons, sites, and varieties can have an effect on the composition of the alfalfa phyllosphere bacterial community. From a production point of view, Pseudomonas [51], Methylobacterium [52,53], and Bacillus [54] were recognized as having the potential to promote plant growth. It has been found that Bacillus and Pantoea suppress fungal growth and are tolerant to conditions such as low water potential and osmotic stress in a study on the biocontrol potential of phyllospheric microorganisms against Exserohilum turcicum [55].
On the leaves of Arabidopsis thaliana, Sphingomonas inhibit the growth of pathogenic Pseudomonas syringae pv. syringae DC3000 and suppress disease symptoms [56]. These findings suggested that the dominant phyllosphere microorganisms are beneficial to the health and productivity of alfalfa. In the production of silage, Pseudomonas is an unfavorable bacterium that can thrive in an anaerobic environment and therefore is harmful for silage production. The biogenic amine produced by Pseudomonas leads to a decrease in protein content, which leads to a decrease in the nutritional quality of the feed [57]. Additionally, Pseudomonas is the primary group of foodborne and hazardous bacteria linked to fresh produce [58]. In our study, Pseudomonas showed higher relative abundance in summer and was positively correlated with temperature and sunshine duration. Therefore, we speculated that alfalfa harvested in the hot and rainy summer was not conducive to silage fermentation. Notably, the relative abundance of bacteria favoring silage fermentation, such as Lactococcus, Lactobacillus, and Weissella, accounted for less than 1% in all our samples. Lactobacillales (Lactobacillus, Lactococcus, and Weissella) are the main heterofermenters in the silage process, inhibiting negative microorganisms and promoting fermentation through the production of metabolites such as organic acids, bacteriocins, and competition for nutrients [59]. Our results show that the alfalfa phyllosphere lacks beneficial bacteria favoring fermentation. Therefore, microbial additives should be used to promote the fermentation of alfalfa before silage.
The phyllospheres of the two sites and varieties of alfalfa showed some degree of overlap in spring and summer and some degree of overlap in autumn and winter ( Figure 3). This indicates that the distribution and structure of bacterial communities in alfalfa differed less in these two periods. The overlap may have been due to the small changes in environmental factors such as temperature, sunshine duration, precipitation, and relative humidity.

Influence of Variety, Site, and Season on the Structure of Alfalfa Phyllosphere Bacterial Communities
Abiotic factors such as environmental conditions (solar UV radiation, temperature fluctuations, relative humidity, and nutrient availability), agronomic practices, and biotic factors such as host plant species all influence the structure of the phyllosphere bacterial community [3,[60][61][62]. Using 16S rRNA gene sequencing, this study investigated the effects of alfalfa variety, plant location, and season on the composition and diversity of alfalfa phyllosphere bacterial communities. When compared to site and variety, season was the most important driver of the phyllosphere bacterial community structure in alfalfa (R 2 = 27.6%). This result is similar to previous findings that show that annual changes have the strongest effect on bacterial communities [63]. The composition of the tree phyllosphere can vary considerably over longer time periods, with temporal variation on a given tree exceeding the variation in community composition among individual trees sampled on a given day, and can exhibit certain seasonal patterns [64]. Although there are other studies reporting different plant materials with different phyllosphere microbiota [22,65], little variation in bacterial community composition has been observed in different species of plants compared to other plant types [18,23]. According to Kim et al. [66], the phylogeny of the host plant determines how similar the phyllosphere bacterial communities of these tree species are to one another, and more similar communities are found in closely related host plants. The selection impact of host plants reduces bacterial diversity, according to Xiong et al. [45]. The bacterial populations in the tobacco phyllosphere have also shown significant resistance to environmental perturbations, according to Chen et al. [67]. These results might suggest that the same host plant exhibits similar bacterial community structures, in which case the impact of different alfalfa types on the bacterial community structure would not be very noteworthy. We also found that the interaction of season and site had the greatest influence on the community structure of alfalfa phyllosphere bacteria (R 2 = 42.9%). However, neither variety, nor the interactions of site and variety, nor the interactions of season, site, and variety significantly impacted the richness and community structure of phyllosphere bacteria. Numerous studies have documented the clear impact of distance on bacterial diversity, which is primarily attributable to site-specific elements such as nutrient availability, temperature swings, relative humidity, and sun radiation [45]. Additionally, Redford et al. [68] hypothesized that each type of tree has a unique structural composition of phyllosphere bacterial population that remains constant even after the trees are planted in various locations throughout the globe. This may help to explain why, in our investigation, the interaction between cultivation sites and variety had no appreciable impact on the diversity and community structure of the phyllosphere.

Correlation Analysis between Bacterial Communities and Environmental Factors
Season and location interactions significantly affect alfalfa surface microorganisms, and this effect may be caused by changes in sunshine duration, temperature, and precipitation [69]. In this study, Sphingomonas, Pseudomonas, and Pantoea were positively correlated with temperature, precipitation, and sunshine duration (p < 0.01), and the dominant bacterial genera on alfalfa samples were Sphingomonas and Pseudomonas. This indicated that the main factors affecting the microbial diversity of the alfalfa phyllosphere were temperature, precipitation, and sunshine duration. In addition, the three aforementioned genera were positively correlated with sunshine duration but negatively correlated with relative humidity. This indicates that they are suitable for growth in hot and dry environments. This result is different from those of Guan et al. [70], who suggested that Sphingomonas is suitable for growth in cool and humid environments. Proteobacteria comprises diazotrophic bacteria that can use atmospheric dinitrogen as a source of nitrogen [71]. The richness of N 2 -fixing bacteria increased with drought, suggesting that their supplementation may expand plant adaptation to the environment [72]. Sphingomonas [73], Pseudomonas [74], and Pantoea [75] all belong to Proteobacteria and have a nitrogen-fixing role, which may explain why these bacterial genera were negatively correlated with relative humidity in our study. Temperature, precipitation, and sunshine duration were negatively correlated with Bacillus, which explains why Bacillus was more abundant in winter than in other seasons. Bacillus are the main spoilage bacteria in silage, mainly degrading proteins and amino acids, and play an important role in silage spoilage deterioration under aerobic conditions [76]. Our research revealed that the bacterial community of the phyllosphere varied, possibly as a result of growing season disturbances brought on by UV exposure, moisture levels, resource availability, or changes in leaf cuticle characteristics. As a result, a "climax" community was not formed, and neither bacterial diversity nor community composition stabilized over time [77].

Conclusions
In this study, we described the bacterial communities found in the phyllosphere of alfalfa and investigated the effects of sample site, season, and variety on phyllosphere community structure. Our main conclusions are that season-site interaction, season-variety interaction, and season accounted for the majority of the variation in phyllosphere bacterial community formation. The environmental factors that contributed most to the phyllosphere bacterial community of alfalfa were temperature and sunshine duration, which were significantly positively correlated with most of the dominant bacterial genera in the alfalfa phyllosphere. The bacterial community in the phyllosphere of alfalfa shows a certain seasonal change over time and high sensitivity to changing environmental parameters. In addition, the alfalfa phyllosphere lacks beneficial bacteria that can initiate silage fermentation. Thus, the addition of exogenous substances to alfalfa silage is recommended to promote its fermentation.

Data Availability Statement:
The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Conflicts of 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.