Pectin Degradation is an Important Determinant for Alfalfa Silage Fermentation through the Rescheduling of the Bacterial Community.

This study aimed to evaluate the effects of the four kinds of additives on the silage quality and the relevant bacterial community diversity by Illumina HiSeq 16S rRNA sequencing. The four kinds of additives were Lactobacillus plantarum (LP), organic acids including gallic acid (GA) and phenyllactic acid (PA), pectin (PEC), and enzymes including pectinase (PEE) and cellulase (CE). After 30 d of fermentation, the pH value was shown to have the lowest value in the PEE and PEC groups, followed by the PA group, and then in CE and GA groups; the highest value of pH was found in both LP and control groups. The ammonia nitrogen concentration was lower in the PEE group compared to the other groups except for the PA group. In the comparisons among the seven groups, Lactobacillus was higher in the LP group, Paracoccus was higher in the GA group, Weissella was higher in the PA group, Leuconostoc was higher in the PEC group, Bacillus, Aeromonas, and Curvibacter were higher in the PEE group, and Coriobacteriaceae_UCG_002 was higher in the CE group compared to the other groups. This study proposed that the addition of PEC and PEE improved the fermentation quality of alfalfa silage compared to other additives by improving the bacterial community of Leuconostoc, and Bacillus and Aeromonas, respectively. Moreover, the enhanced fermentation quality of alfalfa silage by the supplementation of PEC and PEE might be attributed to other unclassified genera. This study provides an implication that pectin degradation is an important determinant for alfalfa silage fermentation through the rescheduling of bacterial community diversity.


Introduction
Silage is one of the main ingredients and energy and digestible fiber sources in ruminant diets, especially for dairy cattle [1]. Silage making is a bioprocessing ideal method for anaerobic preservation and pretreatment of forage, as it can prolong the stable storage of forage and keep the succulence of grass [1,2]. Alfalfa (Medicago sativa L.) is widely used as a forage source in ruminants, but it is difficult to generate high-quality alfalfa silage due to its high buffer capacity and low fermentable carbohydrate [3,4]. Poorly made silage can harbor pathogens, reduce animal performance, cause diseases of cattle, and even potentially endanger human health [5]. It is, therefore, crucial to perform proper quality control by additives' supplementation including organic acids, lactic acid bacteria (LAB), sugars, and enzymes to generate more lactic acid and lower pH value during ensiling [4].
Alfalfa has a relatively high pectin concentration that ranges from 10.5% to 14.2% (dry matter basis) [6]. It has been reported that the addition of pectinase and cellulase can improve the release of

Experimental Design
The experiment was conducted in June, 2019 at one rural land field (Nanyang, Henan, China). Alfalfa was planted in fields of loam soil, and no herbicides or fertilizers were applied. Alfalfa was harvested at the squaring stage after the first cutting. After cutting and wilting for several hours, the alfalfa was chopped into 1-to-2-cm segments with a forage cutter (Lingong Machinery, Shandong, China). The dry matter of the alfalfa material before ensiling was 27.3%. The chopped forage was sampled and analyzed for chemical composition. The concentrations of crude protein, neutral detergent fiber, acid detergent fiber, and WSC in alfalfa before ensiling were 17.1%, 32.8%, 23.4%, and 4.77% of the dry matter basis, respectively. The seven treatments, including a control (CON), Lactobacillus plantarum (LP, 1 × 10 6 cfu/g of fresh matter), gallic acid (GA, 0.5% of fresh matter), phenyllactic acid (PA, 0.5% of fresh matter), pectin (PEC, 65% purity, from citrus, 2% of the fresh matter), pectinase (PEE, 50 U/g, from Aspergillus niger, 0.05% of fresh matter), and cellulase (CE, 10,000 U/g, from Trichoderma reesei, 0.05% of fresh matter), were used for silage and evaluated with three replicates each of alfalfa before ensiling. The LAB strain was isolated from alfalfa (Beijing, China) silage and identified as Lactobacillus plantarum (GenBank accession number: WCFS1) by 16S rRNA gene sequencing [16]. The strain was inoculated into liquid optimized medium for LAB and cultivated for 20 h. Then, the L. plantarum cells were collected by centrifugation at 5000 g for 10 min at 4 • C. The bacterial powder was obtained by freeze drying the cell pellets mixed with non-fat dry milk. Before use, the inoculum was prepared by resuspending the powder in deionized water to a concentration of 1 × 10 12 cfu/g. The added dose of pectinase and cellulase was according to a previous study of Nadeau et al. [7]. The additive was mixed with 2 mL of sterile distilled water and sprayed onto the 200 g chopped alfalfa (only 2 mL of sterile distilled water for CON). The mixed material (200 g) was packed into polyethylene oxygen isolation plastic film bags (Hiryu KN type, dimensions: 180 mm × 260 mm; thickness: 0.2 mm; Embossed Food saver bag; Asahikasei, Tokyo, Japan). Then, the bags were sealed with a vacuum sealer (BH950; Matsushita, Tokyo, Japan) and stored at ambient temperature (22-25 • C) in dark conditions for 30 d of ensiling.

Sample Collection and Measurements
The silage sample (20 g) was mixed with distilled water (180 mL), homogenized in a blender for 30 s, and filtered through four layers of cheesecloth [17]. The pH was measured immediately using a pH meter (PHS-3C, INESA Scientific Instrument, Shanghai, China). The concentrations of lactic acid, acetic acid, propionic acid, and butyric acid were measured by high-performance liquid chromatography (HPLC, Shimadzu, Tokyo, Japan). The contents of ammonia nitrogen were determined according to Broderick and Kang [18].

DNA Extraction and Sequencing
The microbial pellet was obtained from silage according to the procedure from a previous study [19]. The alfalfa silage is a kind of fermented product of feed, which is similar to the stool sample. Thus, the microbial DNA was extracted using the E.Z.N.A. stool DNA kit (Omega Biotek, Norcross, GA, U.S.) according to the manufacturer's protocols. The 16S rDNA V3-V4 region of the eukaryotic ribosomal RNA gene was amplified by PCR using primers 341F: CCTACGGGNGGCWGCAG and 806R: GGACTACHVGGGTATCTAAT. Amplicons were extracted from 2% agarose gels and purified using the AxyPrep DNA Gel Extraction Kit (Axygen Biosciences, Union City, CA, USA) according to the manufacturer's instructions and quantified using QuantiFluor-ST (Promega, Madison, WI, USA). Purified amplicons were pooled in equimolar ratios and paired-end sequenced (2 × 250) on an Illumina platform by Hiseq2500 PE250 (Illumina, San Diego, CA, USA) according to the standard protocols. The raw reads were deposited into the NCBI Sequence Read Archive (SRA; http://www.ncbi.nlm.nih.gov/Traces/sra/) database (Accession Number: SRP250821).

Bioinformatic Analysis
Raw data containing adapters or low-quality reads would affect the following assembly and analysis. The raw sequences were selected according to Wang et al. [20]. Paired-end clean reads were merged as raw tags using FLSAH (v 1.2.11) with a minimum overlap of 10 bp and mismatch error rates of 2%. Noisy sequences of raw tags were filtered by the QIIME (V1.9.1) pipeline under specific filtering conditions to obtain clean tags. To perform a reference-based check for chimera, the clean tags were compared against the reference Gold database (r20110519; http://drive5.com/uchime/uchime_download. html) using UCHIME Algorithm (http://www.drive5.com/usearch/manual/uchime_algo.html).
All chimeric tags were removed to obtain effective tags, then the effective tags were clustered into operational taxonomic units (OTUs) of ≥97% similarity using the UPARSE pipeline. The OTUs were extracted according to the minimum abundance. A Venn analysis was performed to identify unique and common OTUs Between the seven groups. The analysis of the taxonomy assignment of representative sequences was performed by a naive Bayesian model using an RDP classifier that was based on the SILVA (version v123) database. Alpha diversity indices including Sobs, Shannon, Simpson, Ace, Chao, and Coverage were calculated in QIIME. The OTU rarefaction curve and rank abundance curves were plotted in QIIME. Welch's t-test and a Wilcoxon rank test were performed for the alpha index comparisons between two groups. Tukey's honestly significant difference (HSD) test and a Kruskal-Wallis H test were performed to compare the alpha indices among groups. For comparisons of the seven groups, the permutational multivariate analysis of variance (PERMANOVA) based on bray-curtis distances were performed using the vegan package in the R programming environment [21,22]. Multivariate statistical analyses, including a principal coordinates analysis (PCoA) of bray-curtis distances, were calculated. The functional genes of the bacterial communities were predicted using Tax4Fun. The non-strict version of LEfSe (Linear discriminant analysis Effect Size) was utilized to select and demonstrate the differentially abundant taxonomies among the groups by coupling a nonparametric factorial Kruskal-Wallis sum-rank test for statistical significance with additional tests assessing biological consistency and the relevance of effects. Bacteria with linear discriminant analysis (LDA) scores greater than 3 were speculated to have a different abundance.

Correlations Analysis
To explore the functional correlation between the rumen bacterial changes and fermentation quality of alfalfa silage, a Spearman's rank correlation matrix was generated by calculating the Spearman's correlation coefficient among the taxa that were affected by the diet type (at the genus level, p < 0.05) and fermentation characteristics in the R program, and only connections with a p-value of less than 0.01 and r > 0.55 or r < −0.55 were retained. These correlations were visualized using the R program pheatmap package. To detect the relationship of fermentation organic acids with samples, redundancy analysis (RDA) was performed at the genus level using the R program (https://www.r-project.org) vegan package, with organic acid proportions considered as explanatory variables.

Statistical Analysis
The fermentation characteristics were analyzed using one-way Analysis of Variance (ANOVA) analysis and Duncan's multiple range tests based on the GLM procedure of SAS (version 9.2, SAS Institute Inc., Cary, NC, USA). Significance was declared at p < 0.01.

Fermentation Characteristics of Silage
The PEE and PEC groups had the lowest pH value, followed by the PA group, and then by the CE and GA groups, and the highest pH was found in the LP and CON groups ( Table 1). The ammonia nitrogen concentration was lower in the PEE group compared to the other groups except for the PA group. The ammonia nitrogen concentration of CON was in the highest value compared to other groups, followed by LP, CE, and GA. The PEC group had the highest lactic acid concentration and the CON had the lowest value. The acetate proportion was greater in the LP and GA groups compared to the PA, PEC, and PEE groups. The propionate concentration was higher in GA compared to other groups except for PA. a-e Means within a row with different superscripts differ (p < 0.01). 1

Bacterial Profiles in the Silage
In total, 2,025,710 raw reads were obtained for the bacterial 16S rRNA genes in the seven groups. After screening, 1,866,479 effective tags were obtained, accounting for 92.15% of the raw reads. The rarefaction curves showed clear asymptotes, and the number of reads was the same for all samples after normalization ( Figure 1A). The Good's coverage value for all samples was greater than 99.1%. The results of the PCA and PCoA with bray distances indicated that the treatment groups of PA, PEC, and PEE were largely separated from other groups including CON, LP, GA, and CE at the OTU level ( Figure 1B,C). There were 1217 OTUs that were identified in the silage of the seven groups, among which 333 OTUs were found in all the seven groups and accounted for 27.4% of the total OTUs ( Figure 2). There were 76, 121, 59, 173, 107, 293, and 55 OTUs that were identified individually in CON, LP, GA, PA, PEC, PEE, and CE, respectively ( Figure 2). For the alpha diversity analysis, a similar level of species richness existed among the seven groups based on the Sobs, Simpson, Ace, and Chao index analyses, which indicated that there was a similar tendency of diversity and uniformity among the seven groups ( Table 2). The Shannon index analysis indicated that there were diversity and uniformity between CON and GA, between CON and PEC, and between PEC and PEE.  uniformity among the seven groups ( Table 2). The Shannon index analysis indicated that there were diversity and uniformity between CON and GA, between CON and PEC, and between PEC and PEE.
The non-strict version of LEfSe was used to determine the bacteria most likely to explain the differences among the different additives' treatments ( Figure 4). Lactobacillus was higher in the LP group, Paracoccus was higher in the GA group, Weissella was higher in the PA group, Leuconostoc was higher in the PEC group, Bacillus, Aeromonas, and Curvibacter were higher in the PEE group, and Coriobacteriaceae_UCG_002 was higher in the CE group compared to the other groups. The non-strict version of LEfSe was used to determine the bacteria most likely to explain the differences among the different additives' treatments ( Figure 4). Lactobacillus was higher in the LP group, Paracoccus was higher in the GA group, Weissella was higher in the PA group, Leuconostoc was higher in the PEC group, Bacillus, Aeromonas, and Curvibacter were higher in the PEE group, and Coriobacteriaceae_UCG_002 was higher in the CE group compared to the other groups.  It shows a representation of the differentially abundant bacteria (highlighted by small circles and by shading) among the six additives groups (B). There are six layers from the inside of this plot to the outside, corresponding to six levels of the taxonomy (kingdom, phylum, class, order, family, and genus). The size of each node represents their relative abundance.

Correlation Analysis
RDA revealed that the bacterial community structure was affected by fermentation characteristics (including pH, lactic acid, acetic acid, propionic acid, and ammonia nitrogen), in which the length and angle of each arrow represents its degree and direction of correlation to the bacterial community in the RDA (no permutations), respectively ( Figure 5A). As shown in Figures 5A, pH (p = 0.001), lactic acid (p = 0.004), acetic acid (p = 0.001), propionic acid (p = 0.016), and ammonia nitrogen (p = 0.001) significantly affected the bacterial community structure. The samples that belong to the PEE and PEC groups, located in the positive direction of the lactic acid arrow, were positively related to lactic acid, whereas most of the other samples were negatively related to lactic acid.

Correlation Analysis
RDA revealed that the bacterial community structure was affected by fermentation characteristics (including pH, lactic acid, acetic acid, propionic acid, and ammonia nitrogen), in which the length and angle of each arrow represents its degree and direction of correlation to the bacterial community in the RDA (no permutations), respectively ( Figure 5A). As shown in Figure 5A, pH (p = 0.001), lactic acid (p = 0.004), acetic acid (p = 0.001), propionic acid (p = 0.016), and ammonia nitrogen (p = 0.001) significantly affected the bacterial community structure. The samples that belong to the PEE and PEC groups, located in the positive direction of the lactic acid arrow, were positively related to lactic acid, whereas most of the other samples were negatively related to lactic acid.

Discussion
Silage bacterial inoculants, chemical additives, and mixed additives are known for their positive effects such as improving fermentation, increasing nutrients recovery, extending aerobic stability, and mitigating the pathogenicity of silage [4]. For the uneasily fermented materials, the most important manner is to improve their fermentation quality. It has been reported that the pectinase or cellulase appears to contribute to the fermentation of alfalfa silage [7]. In our study, the addition of pectin or pectinase modified the fermentation characteristics including pH and lactic acid concentration. However, the bacterial function underlying the contribution of additives to ensiling alfalfa still needs to be clarified. From the subsequent bacterial sequence data, we found the diversity of the bacterial community contributed to the efficient roles of these additives.
The dominant strains in the alfalfa silage belong to the genera Lactobacillus, Enterococcus, Weissella, and Streptococcus in previous studies of alfalfa silage [23,24], which was a little different from the present study. The dominant three genera bacteria across the seven groups were Lactobacillus, Weissella, and Enterobater. The LP belongs to the genus Lactobacillus, and the significantly greater level of Lactobacillus in the LP group confirmed the supplementation and colonization of LP in alfalfa silage. However, in the current study, the Lactobacillus was positively correlated with the pH value and ammonia nitrogen, which was contrary to the general common findings. We estimated that the contribution of LP in decreasing pH and ammonia nitrogen might be less significant than other silage bacteria. From the view of genera classification, more unclassified bacteria were found in the PEC and PEE groups (Figure 3), which might contribute to the improved fermentation qualities. Thus, we proposed that the contribution of decreasing pH value of alfalfa silage might be due to other key microorganisms, except for Lactobacillus during ensiling.
The genus Weissella, which belongs to LAB, is assigned to the phylum Firmicutes [25], which was greater in the PA group compared to other groups. It was reported that the strains of Weissella cibaria FMF4B16 belong to the genera Weissella and have strong inhibitory activity against food molds [26]. Then, the antifungal compounds such as phenyllactic acid, 2-hydroxy-4-methylpentanoic acid, and other organic acids were investigated to be responsible for the antifungal activity of Weissella. Thus, there existed linkage between phenyllactic acid and Weissella, and we proposed that the phenyllactic acid might contribute to the growth of Weissella, which need more research to clarify it. In addition,

Discussion
Silage bacterial inoculants, chemical additives, and mixed additives are known for their positive effects such as improving fermentation, increasing nutrients recovery, extending aerobic stability, and mitigating the pathogenicity of silage [4]. For the uneasily fermented materials, the most important manner is to improve their fermentation quality. It has been reported that the pectinase or cellulase appears to contribute to the fermentation of alfalfa silage [7]. In our study, the addition of pectin or pectinase modified the fermentation characteristics including pH and lactic acid concentration. However, the bacterial function underlying the contribution of additives to ensiling alfalfa still needs to be clarified. From the subsequent bacterial sequence data, we found the diversity of the bacterial community contributed to the efficient roles of these additives.
The dominant strains in the alfalfa silage belong to the genera Lactobacillus, Enterococcus, Weissella, and Streptococcus in previous studies of alfalfa silage [23,24], which was a little different from the present study. The dominant three genera bacteria across the seven groups were Lactobacillus, Weissella, and Enterobater. The LP belongs to the genus Lactobacillus, and the significantly greater level of Lactobacillus in the LP group confirmed the supplementation and colonization of LP in alfalfa silage. However, in the current study, the Lactobacillus was positively correlated with the pH value and ammonia nitrogen, which was contrary to the general common findings. We estimated that the contribution of LP in decreasing pH and ammonia nitrogen might be less significant than other silage bacteria. From the view of genera classification, more unclassified bacteria were found in the PEC and PEE groups (Figure 3), which might contribute to the improved fermentation qualities. Thus, we proposed that the contribution of decreasing pH value of alfalfa silage might be due to other key microorganisms, except for Lactobacillus during ensiling.
The genus Weissella, which belongs to LAB, is assigned to the phylum Firmicutes [25], which was greater in the PA group compared to other groups. It was reported that the strains of Weissella cibaria FMF4B16 belong to the genera Weissella and have strong inhibitory activity against food molds [26]. Then, the antifungal compounds such as phenyllactic acid, 2-hydroxy-4-methylpentanoic acid, and other organic acids were investigated to be responsible for the antifungal activity of Weissella. Thus, there existed linkage between phenyllactic acid and Weissella, and we proposed that the phenyllactic acid might contribute to the growth of Weissella, which need more research to clarify it. In addition, it was found that the dominant genera Weissella was found in the more easily fermented forages such as corn, sorghum, and forage paddy rice silages [27]. Thus, the dominant and increased abundance of genera Weissella in the PA group might indicate the well-fermented alfalfa silage.
The genus Paracoccus is facultative aerobic, Gram-stain-negative, short rod-shaped, catalase-positive, and oxidase-positive bacteria belonging to the class Alphaproteobacteria. Members of the genus Paracoccus can exist in various environments such as air [28], freshwater [29], and agricultural field [30]. It was reported that one species of Paracoccus, Paracoccus jeotgali la sp. nov., can utilize pectin, acetoacetic acid, myoinositol, propionic acid, glycyl-L-proline, α-keto-glutaric acid, citric acid, L-malic acid, and butyric acid [31]. In addition, it was reported that gallic acid has hydrogen-scavenging activity [32]. Borges et al. [33] reported that gallic acid induced irreversible changes in E. coli, P. aeruginosa, S. aureus, and L. monocytogenes membrane properties by the interaction of gallic acid with bacterial cytoplasmic membrane causing hydrophobicity changes, a decrease in negative surface charge, and local rupture and pore formation with leakage of intracellular constituents. Thus, the results of the current study may offer a new finding: that the addition of gallic acid has the ability to induce the accumulation of Paracoccus by utilizing gallic acid.
Leuconostoc, positively correlated with lactic acid, was the dominant genera in the PEC group alfalfa silage [27]. It was found that Leuconostoc was the major bacterial genus present from the initial to the middle stages of fermentation [34]. All the bacteria belonging to Leuconostoc could utilize D-fructose [35]. Leuconostoc was enhanced by the supplementation of pectin in the current study, which might indicate that the pectin is an important substrate for the growth of Leuconostoc during alfalfa ensiling.
Bacillus species can produce compounds displaying antifungal activity [36]. Bacillus amyloliquefaciens has been reported to be a promising candidate for new pharmaceutical agents and probiotics [37]. Bacillus has been proven to produce bacteriocins, bacteriocin-like substances, and lipopeptides to exert antifungal effects [38]. Bacillus subtilis SPB1 lipopeptides exhibit surfactant properties and antifungal activity [39]. Bacillus velezensis was evaluated as the antagonistic properties towards toxigenic molds in silage conditions [40], indicating the bioactive roles of PEE in promoting the growth of Bacillus to exert the antifungal properties. Aeromonas was evaluated for pseudomonads fermenting carbohydrates with the production of carbon dioxide and hydrogen [41]. It was reported that Aeromonas caviae is potentially vital for acidogenesis in anaerobic digestion, and can be grown on glucose [42]. Thus, the increased abundance of Aeromonas in the PEE group might be due to the release of carbohydrates such as pectin or other easily fermented sugars by the pectinase. However, we did not explain the linkage between Curvibacter and pectinase as well as the contribution of Curvibacter to the fermentation of alfalfa silage.
The results from the present study support the hypothesis of adding specific additives can improve alfalfa silage fermentation by manipulating the bacterial community. However, we believe that the limitation of this study was mainly due to the relatively low number of samples. The function of the differential genera bacteria should be studied for their targeted efficient strains, such as the growth of Bacillus and Aeromonas in silage. Moreover, we still found many unclassified genera bacteria in the PEC and PEE groups. Given the existence of methods to manipulate the inoculants, the sequencing data should be used for further screening and enrichment culture [43]. Manipulating the levels of species or even strains belonging to the Weissella, Leuconostoc, Bacillus, and Aeromonas, as well as other unclassified bacteria that potentially have a close connection to the anaerobic fermentation of alfalfa silage, which may help to improve the quality of alfalfa silage and the development of associated feed industry.

Conclusions
Overall, the PEC and PEE groups had more efficient roles in contributing to the fermentation of alfalfa silage by improving lactic acid concentration and decreasing pH value and ammonia nitrogen proportion compared to the CE, GA, PA, and LP groups. It was proposed that the addition of PEC and PEE improved the fermentation quality of alfalfa silage by improving the growth of Leuconostoc, and Bacillus and Aeromonas, respectively. The differential bacterial community and unclassified bacteria provided new information on screening targeted or unknown functional bacterial inoculants for modulating alfalfa silage quality.