Skip to main content
Chemical and Biological Technologies in Agriculture
Download PDF
Download ePub

Ensilage using Leuconostoc lactis and Weissella confusa reduces microbial risk and enhances hygienic quality of whole-crop corn

Chemical and Biological Technologies in Agriculture volume 11, Article number: 44 (2024) Cite this article

Abstract

This study combined applied PICRUSt2 and BugBase tools to evaluate the effects of these two strains on the fermentation characteristics, microbial community, potential microbial risk and hygienic quality of whole-crop corn (WCC) silage. Fresh WCC harvested at the dough stage was ensiled with distilled water (CON), Leuconostoc lactis (LS) and Weissella confusa (WA) for 2, 4, 8, 15 and 30 days. After ensiling, all WCC silages presented desirable fermentation with high lactic acid and Lactobacillus proportions, low pH and ammonia nitrogen levels and absent butyric acid. Ensiling decreased the complexity of bacterial co-occurrence networks, and the Lc. lactis and W. confusa inoculation further decreased the complexity. The inoculation of W. confusa suppressed the most pathogenic pathways and related modules associated with zoonosis. In bacterial phenotype predicted analysis, although CON had lower proportions of ‘Potentially pathogenic’ than fresh material, this undesirable phenotype declined to negligible levels via LS and WA inoculation. Even for well-fermented WCC silages, the risk of pathogens remained after 30 days of ensiling. Therefore, WA could be developed as a promising fast start-up inoculant for reducing pathogenic contamination and improving hygienic quality of silage.

Graphical Abstract

Background

Whole-crop corn (WCC) silage is a commonly utilized roughage in ruminant feeding systems, particularly for dairy cattle, owing to its high energy content and year-round availability [1, 2]. The quality and safety of WCC is the key to the production of green and healthy animal products. Our previous study found that, even if silage fermentation was good, harmful microorganisms such as Enterobacter appeared again in large abundance at the late stage of ensiling due to the existence of some acid-resistant strains [3, 4]. Since the proposal of silage technology, additives have played an important role in improving the quality and ensuring the safety of silage [5]. Among various additives, lactic acid bacteria (LAB) inoculants are widely promoted and used because of their safe, stable, effective, low-cost and pollution-free properties. At present, the most studied antibacterial inoculant is Lactobacillus buchneri, which has been reported to require 50 days or longer to function [6, 7]. It is common in practical production that dairy producers have to feed silage after a short ensiling period, and in this case, L. buchneri will not have enough time to produce antibacterial substances such as acetic acid and 1,2-propanediol [8]. Differently, early colonizers in the anaerobic fermentation process, such as heterofermentative Weissella and Leuconostoc, have garnered attention for their potential as fast start-up antibacterial inoculants [9].

Although Leuconostoc lactis and Weissella confusa are relatively new to silage research, they are widely known for their antimicrobial role in the food industry. W. confusa isolated by Tenea and Lara [10] has been found to have obvious inhibitory effects on Escherichia coli, Salmonella and Shigella. And the strains of Lc. lactis also effectively inhibited the growth of harmful microbes including Listeria, Bacillus, Streptophyta, Mycospherella, Aureobasidium and Phoma [11, 12]. Besides, their probiotic properties have been investigated, as evidenced by excellent gastrointestinal tolerance [13], exopolysaccharide-producing capacity [14], adhesion to intestinal epithelial cells [15], and inhibitory activity against intestinal pathogens [12, 16]. These characteristics of W. confusa and Lc. lactis make them promising candidates for inhibiting harmful microbes and enhancing silage quality and overall animal performance. However, to our knowledge, there are no studies on the effects of W. confusa and Lc. lactis on the quality and antimicrobial role of WCC silage.

A comprehensive understanding of the variation of silage microbiota induced by W. confusa or Lc. lactis inoculation will help the safe and hygienic production process of WCC silage. In recent years, culture-independent techniques represented by single-molecule real-time (SMRT) sequencing technology have assisted researchers in analyzing the bacterial community structure and diversity at species-level precision [17]. In addition to the bio-information mentioned above, functional and phenotypic annotation of microbial communities is increasingly found to be extremely important for microbial ecology research. Ward et al. [18] argued that the analysis of microbial communities should take into account phenotype, especially pathogenic potential. The phenotype prediction underlying bacterial communities can be used as an effective supplement to SMRT sequencing analysis and will be highly beneficial for silage research.

Therefore, this study aims to investigate the effects of Lc. lactis and W. confusa as silage inoculants on fermentation characteristics, microbial community dynamics, potential functionality, and hygienic quality in WCC silage.

Methods

Material preparation

Fresh WCC (Zea mays L. Heyu 9566) was cultivated in the Baima experimental station (31°60′N, 119°17′E, mean elevation: 24.9 m, Jiangsu, China). At the dough stage (one-third milk line), fresh WCC was harvested at 8:00 am under clear weather conditions leaving a stubble of 15 cm and cut into lengths of about 30 mm by a fodder cutter. The Lc. lactis CICC 24759 and W. confusa CICC 24453 used in this study were supplied by the China Center of Industrial Culture Collection (CICC) and their inoculant level was 1 × 106 CFU/g fresh weight, respectively. After manual mixing, the chopped material was randomly divided into 3 parts and treated with (i) distilled water (CON), (ii) Lc. lactis CICC 24759 (LS), (iii) W. confusa CICC 24453 (WA). The applied volume of distilled water or inoculants was 1 mL per 100 g fresh weight. Thereafter, about 0.45 kg of the above material was loaded into a laboratory-scale silo (polythene plastic bag with a size of 30 × 40 cm). A total of 45 laboratory-scale silos (3 treatments × 5 storage lengths × 3 replicates) were prepared, vacuum-sealed and stored at the surrounding temperature (27.5 ± 2.5 °C). On day 2, 4, 8, 15 and 30 of ensilage, triplicate bags per treatment were sampled and mixed completely for the following analyses.

Chemical composition, microbial number and fermentation parameter analyses

Chemical composition and microbial number were assessed using the analytical methods outlined in our prior work [4]. For fermentation quality analysis, raw or fermented material was extracted with deionized water at a 3:1 ratio under 4 °C for 30 min. Thereafter, the pH value was determined by a pH meter after filtration with 4 layers of cheesecloth and filter paper. Ammonia nitrogen concentration was determined using colorimetry after reacting with phenol and hypochloric acid [19] in 96-well plates (H803-96, BDBIO, Hangzhou, China) under thermostatic water bath (DWB20-S, DLAB Scientific Co., Ltd., Beijing, China). Fermentation products including lactic, acetic, propionic, butyric acid, ethanol and 1,2-propanediol of fermented material were determined by an Infinity 1260 HPLC (Agilent Technologies Inc., California, USA) equipped with a Carbomix® H-NP5 column (Sepax Technologies, Inc., Delaware, USA) under the following condition: mobile phase of 2.5 mM H2SO4 at a flow rate of 0.5 mL/min and a column run temperature of 55 °C. HPLC internal standard L-lactic acid was purchased from Solarbio Science & Technology Co., LTD. (Beijing, China) and others from Sigma-Aldrich (Shanghai, China).

SMRT sequencing and data analyses

After homogenizing with autoclaved NaCl solution (0.85%) at a ratio of 1:9 and shaken at 37 °C and 120 rpm for 2 min. The above solution was 4 °C centrifuged at 10,000g for 15 min to collect bacterial precipitates. Then, bacterial DNA was extracted from bacterial precipitates using the Rapid Bacterial Genomic DNA Isolation Kits (B518225, Sangon Biotech Co., Ltd., Shanghai, China) and tested by 1% agarose gel electrophoresis. The DNA concentration and integrity were checked utilizing the NanoDrop® 2000 ultraviolet–visible spectrophotometer (260/280 nm, Thermo Scientific, Delaware, USA). Based on primers 27F and 1492R [2], the near full length of the bacterial 16S rRNA gene was amplified utilizing the ABI GeneAmp® 9700 PCR amplification instrument (Applied Biosystems, California, USA) with the running program of 95 °C for 3 min, 27 cycles of 95 °C for 30 s, 60 °C for 30 s and 72 °C for 45 s, with a final extension of 72 °C for 10 min. These amplification products were firstly recovered utilizing 2% agarose gel, then purified by AxyPrep® DNA Gel Extraction Kit (Axygen Biotechnology Co., Ltd., California, USA) and lastly checked by 2% agarose gel electrophoresis. Based on the initial qualitative result of electrophoresis, the purified PCR products were assessed quantitatively using the Quantus™ Fluorometer (Promega Biotechnology Co., Ltd., Wisconsin, USA), in accordance with the sequencing specifications. After library construction using SMRTbell® Express Template Prep Kit 2.0, the purified amplicon was subjected to the Pacbio Sequel II System (Pacific Biosciences of California, Inc, California, USA) for SMRT sequencing.

After receiving the PacBio offline data of subreads, the circular consensus sequencing (CCS) sequence was obtained by SMRTLink 11 software. Subsequently, according to the barcode, the data of each sample were distinguished and the length was filtered. UPARSE (ver. 7.0.0) was used to generate operational taxonomic units (OTUs) through the cluster of the sequences at a 97% identity threshold, and UCHIME (ver. 4.1) was used to identify and remove chimeric sequences. The taxonomic classification of the OTUs was determined by matching against the SILVA database (ver. 138) under the comparison threshold of 70% using the Ribosomal Database Project (RDP) classifier (ver. 2.11). Before bioinformatic analyses, the sequencing data were normalized to the same number of reads per library to avoid sampling bias effects. In brief, sequencing flattening was conducted according to the minimum number of sample sequences and the number of valid sequences for each sample used for analysis is 20,176. QIIME (ver. 1.9.1) was used to calculate the bacterial α-diversities (Shannon, Chao, Ace, Sobs, Simpson and Coverage indices) and β-diversities presented by the Bray–Curtis metric. Bacterial β-diversities were visualized via principal coordinate analysis (PCoA) plots, which were constructed using the vegan package (ver. 2.6–2) of R software (ver. 4.1.2).

Bacterial co-occurrence network analysis

The Networkx toolkit (ver. 2.6.3, https://networkx.org/) was utilized to calculate multiple abundance correlations and establish the co-occurrence pattern among bacterial taxa based on a matrix at the species level [20]. Only species with a relative abundance of 0.05% were considered. A co-occurrence was considered robust if the coefficient of Spearman’s correlation (ρ) was greater than 0.5 and the P-value was less than 0.05. Although a wide range of methods and algorithms are available to construct microbial networks, starting with basic Spearman rank correlation-based approaches, it should be noted that the choice of a correlation metric that is insensitive to compositional effects (Spearman) may lead to the identification of spurious correlations [21]. This is a limitation of most current works and need to be addressed in the future. The Gephi interactive platform (ver. 0.9.2) was used to visualize the co-occurrence networks, with the node representing the bacterial species and the edge representing the correlation between nodes [22]. The topological features of co-occurrence networks, including node and edge numbers, negative and positive correlation numbers, etc., were calculated. The identification of keystone taxa was based on a composite score of low betweenness centrality, high closeness centrality and high mean degree [23].

Bacterial functionality and phenotype prediction analyses

The PICRUSt2 tool developed by Douglas et al. [24] and the BugBase tool (https://bugbase.cs.umn.edu/index.html) [18] were utilized to predict KEGG functional profiles and phenotypes of bacterial communities, respectively. In PICRUSt2, the characteristic sequence (16S rRNA) was aligned with the reference sequence of the Integrated Microbial Genomes (IMG) to construct a phylogenetic tree, find the ‘nearest species’ of the characteristic sequence, and predict the gene information of unknown species according to the gene type and abundance information of known species, so as to predict the pathway of the whole community in combination with the KEGG pathway information of the gene. Compared with PICRUSt, PICRUSt2 can direct predict functionalities for any 16S sequences from OTUs or ASV by taking a sequence placement approach. Default database of reference genomes used for prediction is based on 41,926 bacterial and archaeal genomes from the IMG database, which has been expanded more than 10 times. PICRUSt2 predictions based on several gene family databases including KEGG database, and inference of pathway abundances now relies on MinPath (ver. 1.4), which makes these predictions more stringent. BugBase first normalized the OTU by the predicted 16S rRNA copy number, and then predicted the microbial phenotype using the pre-calculated files provided. BugBase algorithm relies on databases such as IMG, KEGG, and the Pathosystems Resource Integration Center (PATRIC) to predict phenotypes and corresponding microbial contributors at the genus level. The predicted phenotypes include seven categories: Gram positive, Gram negative, biofilm forming, potentially pathogenic, mobile element containing, oxygen utilizing (including aerobic, anaerobic, and facultatively anaerobic), and stress tolerant.

Statistical analysis

The GLM in SAS (ver. 9.2; SAS Institute Inc., North Carolina, USA) was applied to evaluate the effect of cutting height, storage length and their interactions on chemical compositions, fermentative parameters and microbial numbers of fermented WCC. Meanwhile, the Student’s t-test, one-way ANOVA, ANOSIM with 999, Student’s t-test and one-way ANOVA permutations and Wilcoxon rank-sum test were employed to determine whether differences occurred between cutting heights in the characteristics of raw material, α-diversity and β-diversity indices, KEGG functional profiles as well as phenotypes of bacterial communities, respectively. P < 0.05 was deemed significant.

Results

Chemical and microbial compositions of fresh WCC

The pre-ensiled properties of WCC are shown in Table 1. Fresh WCC had a high water soluble carbohydrate content of 257 g/kg dry matter, moderate dry matter content of 298 g/kg FM and relatively low crude protein content of 75.9 g/kg dry matter. The count of epiphytic LAB was 6.34 log10 CFU/g fresh weight, which was more than 5.0 log10 CFU/g fresh weight, and the counts of aerobic bacteria, yeasts, molds and enterobacteria were 8.00, 4.10, 2.96 and 5.14, respectively.

Table 1 The chemical composition and microbial numbers of fresh WCC
Full size table

Chemical and fermentation characteristics of WCC silage

Inoculant had significant (P < 0.001) impacts on the pH value, the water soluble carbohydrate content, the concentrations of acetic acid, propionic acid, ethanol and ammonia nitrogen, and the ratio of lactic acid to acetic acid (LA/AA), and fermentation day had significant (P < 0.01) impacts on the pH value, the content of dry matter and water-soluble carbohydrate, the concentrations of lactic acid, acetic acid, propionic acid, ethanol and ammonia nitrogen as well as the LA/AA of WCC silage (Table 2). The pH value of all treatments decreased sharply within the first 2 days of ensiling and remained at a low level after 4 days of ensiling. The lowest pH value was observed in WA, followed by LS and finally CON. Along with ensiling, the content of dry matter and water-soluble carbohydrate ceaselessly (P < 0.01) decreased, while the concentrations of lactic acid, acetic acid, propionic acid, ethanol and ammonia nitrogen continuously (P < 0.001) increased. Although there were no significant (P > 0.05) differences for dry matter content and lactic acid concentration among treatments, LS and WA showed higher (P < 0.001) acetic acid concentration and water-soluble carbohydrate content but lower (P < 0.001) concentrations of ethanol and ammonia nitrogen than CON. Correspondingly, a higher (P < 0.001) LA/AA value was observed in CON than that in LS and WA silages. Except for 1,2-propanediol detected in LS on day 60, no butyric acid or 1,2-propanediol was found in all WCC silages.

Table 2 Effects of inoculant on fermentation characteristic of WCC silage during ensiling
Full size table

Bacterial community of fresh WCC and WCC silage

The rarefaction curves based on observed OTUs and Shannon index reached near saturation as the number of reads sampled increased (Fig. 1A). The bacterial α-diversity parameters represented by Shannon and Chao indices are shown in Fig. 1B. Compared with FM, except for CON, ensiling process decreased Shannon and Chao indices, and LS and WA inoculation further decreased (P < 0.01) Chao index (Fig. 1B). Each sequenced sample had a Coverage index exceeding 0.98 (Fig. 1C). A significant difference (ANOSIM, P = 0.001) in bacterial β-diversity is shown by the PCoA plot in Fig. 1D. The sample points of 0-day, 4-day and 30-day fermented WCC were clearly distributed in different quadrants.

Fig. 1
figure 1

Bacterial community diversities of fresh WCC and WCC silage. A Rarefaction curves based on observed the OTUs number and Shannon index. B Alpha diversities (Shannon, Chao) of the bacterial community. C Alpha diversities (Coverage indices) of the bacterial community. D Beta diversities of the bacterial community, calculated by principal coordinate analysis (PCoA) based on the Bray–Curtis distance metric. FM, fresh WCC; CON, control (WCC treated with distilled water); LS, WCC inoculated with Lc. lactis, WA, WCC inoculated with W. confusa

Full size image

The bacterial community abundance was presented at the genus and species levels (Fig. 2A, B). Staphylococcus and Mammaliicoccus, unclassified (u.) Staphylococcus and Mammaliicoccus sciuri dominated the bacterial community of raw material at the genus and species level, respectively. After ensiling, the above bacteria were occupied by first cocci LABs and then rod LABs. Concretely, after 4 days of ensiling, Lactococcus lactis significantly (P < 0.001) increased in CON, Lc. lactis significantly (P < 0.01) increased in LS, and W. confusa significantly (P < 0.05) increased in WA (Fig. 2C). After 30 days of ensiling, Lactiplantibacillus plantarum emerged as the prevailing species in all bacterial communities, with relative abundance accounting for 93.1%, 92.9% and 97.2%, respectively. Among them, WA treatment had a significantly (P < 0.001) higher Lactiplantibacillus plantarum than CON and LS treatments.

Fig. 2
figure 2

Bacterial community abundances of fresh WCC and WCC silage. A The bar plot of bacterial community abundance from the genus level. B The bar plot of bacterial community abundance from the species level. C Relative abundance of representative species. FM, fresh WCC; CON, control (WCC treated with distilled water); LS, WCC inoculated with Lc. lactis, WA, WCC inoculated with W. confusa

Full size image

Co-occurrence networks of bacterial community

To investigate the impacts of the ensiling process and inoculants on the relationships between bacterial species, separate 4 bacterial co-occurrence networks were created for fresh and ensiled WCC using Spearman’s rank correlation (Fig. 3A–D). u. Staphylococcus, Devosia aurantiaca, Sphingobacterium populi, etc. in FM, Lactococcus garvieae and Weissella oryzae in CON, Lactococcus lactis, u. Leuconostoc and Levilactobacillus brevis in LS and Lactococcus lactis, Lactococcus garvieae and W. confusa in WA were identified as keystone taxa in the respective co-occurrence networks (Fig. 3E–H). The highest node number was found in the bacterial network of FM, followed by CON, WA, and LS, while the highest edge number was observed in FM, followed by CON, LS, and WA (Fig. 3I). These results suggested that the overall ranking of bacterial networks in terms of nodes and edge number were ranked as follows: FM > CON > LS ≈ WA. The number of positive correlated edge was, respectively, 135, 46, 22, and 15 in FM, CON, LS, and WA and the number of negative correlated edge was, respectively, 16, 18, 21, and 17 in FM, CON, LS, and WA, which implying that except for WA, the correlation edges of all bacterial co-occurrence networks were primarily positive, with negative/positive ratio of 0.12, 0.39, 0.95, and 1.13 in FM, CON, LS, and WA, respectively (Fig. 3J). Ensiling process and WA inoculation had notable (P < 0.05) impacts on the average degree of the bacterial co-occurrence networks (Fig. 3K).

Fig. 3
figure 3

Bacterial co-occurrence networks of fresh WCC and WCC silage. Bacterial co-occurrence networks of FM (A), CON (B), LS (C), and WA (D). Node represents bacterial species, node color represents bacterial phylum, and node size represents relative abundance. Edges are colored according to negative (green) and positive (red) correlations. E–H Elliptic scatter plots of node-level topological features (closeness centrality and mean degree or closeness centrality and betweenness centrality) in four co-occurrence networks. I Bar plots of nude and edge numbers, respectively. J Bar plots of correlation number and negative/positive ratio. K Overall comparison of the mean degree before and after ensiling or among three treatments. The line within the boxplot represents medians, the tops and bottoms of the boxplot represent 25th and 75th percentiles, and lines extend to 1.5 × inter-quartile ranges. The statistical analysis was conducted using a two-sided t-test. P-value is indicated by ***P < 0.001, and by *P < 0.05. FM, fresh WCC; CON, control (WCC treated with distilled water); LS, WCC inoculated with Lc. lactis, WA, WCC inoculated with W. confusa

Full size image

KEGG pathogenic analysis of bacterial community

For KEGG pathway analysis, the KEGG pathogenicity of fresh and ensiled WCC focused on “Infectious disease: bacterial” at pathway level 2, and specifically “Staphylococcus aureus infection”, “Pathogenic Escherichia coli infection” and “Shigellosis” at pathway level 3 (Fig. 4A). The FM had the maximum abundance of “Staphylococcus aureus infection” and “Shigellosis”, followed by CON and LS and finally WA treatment. The “Pathogenic Escherichia coli infection” exhibited a slightly different trend, with the maximum abundance in CON and the minimum relative abundance in WA treatment.

Fig. 4
figure 4

Bacterial potential pathogenic pathways (A) and modules (B and C) of fresh WCC and WCC silage. EHEC/EPEC: enterohemorrhagic E. coli/enteropathogenic E. coli. FM, fresh WCC; CON, control (WCC treated with distilled water); LS, WCC inoculated with Lc. lactis, WA, WCC inoculated with W. confusa

Full size image

For KEGG module analysis, two pathogenic modules (M00542 and M00860) were closely associated with zoonosis. After ensiling, CON treatment enhanced the abundance of “EHEC/EPEC pathogenicity signature (M00542)” but WA treatment declined its abundance (Fig. 4B). As for “Bacillus anthracis pathogenicity signature (M00860)”, although the abundance was still high in CON and LS treatments after 4 days of ensiling, the abundance in all treatments, especially LS and WA, dropped to negligible levels after 30 days of ensiling (Fig. 4C).

BugBase phenotypic analysis of bacterial community

Phenotypic prediction results (Fig. 5A) suggested that the ensiling process significantly (P < 0.05) reduced the abundance of bacterial phenotypes in terms of “Contains Mobile Elements” and “Potentially Pathogenic”. Among all treatments, LS and WA inoculation further (P < 0.05) reduced the abundance of the above phenotypes with the lowest level in WA on day 30. Genera contribution for "Potentially Pathogenic" was visualized in a bar plot (Fig. 5B) and showed that Staphylococcus and Mammaliicoccus, u. Enterobacteriaceae and Enterobacter might be responsible for the pathogenicity risk of fresh material and CON silage, respectively. Overall, a lower (P < 0.001) proportion of “Potentially Pathogenic” was observed in ensiled materials compared to fresh materials, and also in LS and WA treatments compared to CON treatment (Fig. 5C).

Fig. 5
figure 5

Bacterial phenotypes annotation of fresh WCC and WCC silage. A Phenotypes that reflect bacterial characteristics including “Gram Negative”, “Gram Positive” and “Contains Mobile Elements” and bacterial resistance including “Forms Biofilms”, “Stress Tolerant” and “Potentially Pathogenic”. B Bar plot of pathogenic contribution analysis. C Overall comparison of the potentially pathogenic before and after ensiling or among three treatments. The statistical analysis was conducted using a two-sided t-test. P-value is indicated by ***, which represents significant at P < 0.001. FM, fresh WCC; CON, control (WCC treated with distilled water); LS, WCC inoculated with Lc. lactis, WA, WCC inoculated with W. confusa

Full size image

Discussion

The pH values of CON, LS and WA after 4 days of ensiling were always < 4.20, indicating that WCC silages could be effectively preserved even without additives [5]. Such a high fermentation rate could be respectively related to the sufficient LAB number, abundant water-soluble carbohydrate content and low BC of raw material, which makes WCC the most popular feedstock for silage making [25]. Even so, LS and WA could still effectively accelerate lactic acid production during the early phase of ensiling to further reduce pH value, showing the great potential to function as fast start-up inoculants in silage. Similar to most studies of silage fermentation, with the prolongation of ensiling, acetic acid concentration continuously increased. The relatively high acetic acid concentration after inoculation of LS and WA is to be expected because these two strains are heterofermentative LABs, which also explains the lower LA/AA in LS and WA than in CON silages. Different from LS, WA had relatively high LA/AA (> 3.0) during the whole ensiling, indicating that its fermentation pattern was the homofermentative type [6]. This may be because the addition of WA promoted the growth of homofermentative LABs such as L. plantarum, which could also be supported by sequencing data. According to Cai et al. [26], lactic acid-producing cocci initiate lactic acid fermentation and establish a weak-acid environment suitable for lactobacilli. Similar growth-promoting effects on L. plantarum were also found in L. buchneri [27], Bacillus amyloliquefaciens and B. subtilis [2]. The presence of butyric acid in silage fermentation is undesirable. The absence of butyric acid in all WCC silages suggested no clostridia, as these microorganisms are mainly responsible for producing butyric acid during ensiling. Similar to butyric acid, high concentrations of ethanol are also commonly associated with undesirable fermentation [5]. The higher ethanol concentrations in CON than that in LS and WA at the end of ensiling could be associated with the non-negligible yeast and enterobacteria number in CON silage. Certain heterofermentative LABs such as Lentilactobacillus buchneri are known for the production of 1,2-propanediol by metabolizing lactic acid. So, the presence of 1,2-propanediol in LS at the end of ensiling could be associated with the dominant heterofermentation.

The sufficient water-soluble carbohydrate content in raw material might contribute to the high residual water soluble carbohydrate content (> 100 g/kg dry matter) at the end of ensiling. And the higher residual water soluble carbohydrate content in LS and WA could be attributed to their rapid pH decline and high acetic acid production inhibiting the sugar consumption from undesirable microbes. As an indicator of protein degradation and amino acid deamination, the ammonia nitrogen parameter has been widely used to assess the fermentation quality of silage. Ammonia nitrogen concentrations less than 10% (100 g/kg) of total nitrogen indicate good fermentation [28]. The ammonia nitrogen concentrations of all WCC silages were always below the prescribed threshold, showing no extensive proteolysis and amino acid loss. Protein degradation, amino acid decarboxylation and ammonia formation during anaerobic fermentation are attributed to various factors, including plant protease, clostridia, and enterobacteria activities [29, 30]. Given that no clostridia (reflected by butyric acid concentration and microbial community analysis) was detected in all WCC silages, the lower ammonia nitrogen concentration in LS and WA than that in CON was partly due to the relatively low enterobacteria number in LS and WA. Moreover, the rapid acidification in LS and WA to inhibit ammonia nitrogen production caused by plant proteases is also considered.

The rarefaction curves reached near saturation, indicating the sequencing data were sufficient and reasonable. The high Coverage index (> 99%) implied that the majority of the bacterial community in each sample was successfully captured [31]. The LS and WA inoculation reduced the bacterial α-diversity (diversity and richness) in WCC silage, as evidenced by the lower Shannon and Chao indices in LS and WA than in FM and CON, which could be possibly due to their previously validated antibacterial effects [10,11,12]. Meanwhile, extensive studies confirmed that pH decline in microenvironments can reduce bacterial diversity. Similar results were also observed in this study, that is, lower pH was associated with lower bacterial α-diversity.

The dissimilarities in bacterial community composition were visualized through distances between symbols in the PCoA plot. The separation among fresh material, 4-day silages, and 30-day silages suggested the obviously different bacterial community compositions of WCC at the various phases of ensiling (pre-ensiling, early and late). These differences could also be ascribed to the inactivation of acid-sensitive bacteria and the succession of acid-tolerant bacteria while fermenting, as mentioned previously. And the cluster among 4-day silages or 30-day silages showed that no bacterial composition differences were found among different treatments on the same ensiling day.

In this study, the opportunistic pathogens e.g. u. Staphylococcus and Mammaliicoccus sciuri were highly distributed in fresh WCC material, and even u. Brucella and Brucella anthropi were also detected. These results can be interpreted by the inevitable contamination of microorganisms and pathogens from manure, organic fertilizer, irrigation water, air, or soil during plant growth [32]. Similar to most silage studies, the distinct succession from Proteobacteria to Firmicutes before and after fermentation can largely attributed to the acidic microenvironment which suppressed acid-sensitive bacteria (u. Staphylococcus, Mammaliicoccus sciuri, u. Brucella, Brucella, etc.) and promoted LABs. Also, the inoculation of LS and WA respectively increased the proportions of Lc. lactis and W. confusa in the resulting 4-day WCC silages, which was expected. Cocci LABs such as Weissella and Leuconostoc are regarded as the early colonizers in silage fermentation because of their ability to adapt to weak acid environments and rapid acid production [33], While, with the progress of ensiling and further decline of pH in silos, cocci LABs are less tolerant to acid than Lactobacillus. Thus, the moderately acidic environment created by cocci LABs benefits the growth of rod LABs. Correspondingly, at the end of ensiling, the bacterial community of all WCC silages was ultimately occupied by Lactobacillus.

Bacterial co-occurrence networks of fresh WCC or WCC silages were built to comprehensively understand the impacts of LS and WA inoculation on the correlation and interaction of the resulting microbiota. According to node and edge number, the bacterial community complexity of WCC suffered a great reduction after ensiling. The inoculation of LS or WA further reduced the bacterial community complexity resulting in the simplest bacterial correlation structures in WA. These results indicate that (1) ensilage tends to reduce bacterial network complexity; and (2) the antimicrobial properties of LS and WA explain the lower bacterial network complexity.

Banerjee et al. [34] found that the negative correlation in microbial co-occurrence networks suggests a potential competition for resources and shared predators, whereas the positive correlation suggests mutualistic or collaborative associations among microbial taxa. In the current study, the highest positive correlation proportion was observed in the bacterial network of FM. The primarily cooperative nature of the microbiota in fresh WCC implies that the survival of epiphytic bacteria and the stable coexistence of the bacterial community on the plant surface is linked to the collaboration among bacterial cells. It is worth noting that the higher negative correlation proportion in the bacterial network of LS and WA than that in the bacterial network of CON reflects the increased competitive interactions among bacterial species after inoculation. This seems to confirm the competition conjecture between exogenous LABs and herbage epiphytic bacteria proposed by previous studies [35, 36].

According to the method of Berry and Widder [23], u. Staphylococcus, Devosia aurantiaca and Sphingobacterium populi in FM, Lactococcus garvieae and Weissella oryzae in CON, Lactococcus lactis, u. Leuconostoc and Levilactobacillus brevis in LS, and Lactococcus lactis, Lactococcus garvieae and W. confusa in WA were identified as the keystone taxa, correspondingly. The keystone taxa Sphingobacterium populi identified in fresh WCC belongs to Sphingobacteriales, which was consistent with the previous study that Sphingobacteriales was regarded as the keystone taxa in grasslands, forest, or woodlands [37]. Meanwhile, Banerjee et al. [37] found that although keystone taxa have considerable effects on bacterial communities and functions, their abundance is not proportional to their effects. Similarly, the keystone taxa in this work were not necessarily the ones with high relative abundance.

In recent years, emerging tools such as PICRUSt and Tax4Fun have been proposed to assist sequencing analysis for predicting the microbial KEGG functionalities, not only in terms of metabolism but also pathogenicity [24]. Herein, the new PICRUSt2 tool was adopted to reveal the impact of ensilage operation and further LS and WA inoculation on the potential bacterial pathogenicity of WCC silage. Inappropriately managed silage is a source of pathogenic bacteria in ruminants, and feed-borne Staphylococcus aureus, Escherichia coli, Shigella and Bacillus anthracis are responsible for mastitis, hemorrhagic colitis, shigellosis and anthrax in dairy cows, respectively [38, 39]. Thus, three pathogenic pathways of Staphylococcus aureus infection, pathogenic E. coli infection and shigellosis as well as two pathogenic modules of enterohemorrhagic E. coli/enteropathogenic E. coli (EHEC/EPEC) pathogenicity signature and Bacillus anthracis pathogenicity signature are the focus of this study. The spontaneous ensilage process reduced the risk of Staphylococcus aureus infection and shigellosis but not pathogenic E. coli infection, which might be ascribed to the presence of some acid-resistant strains of E. coli. Although most E. coli is considered not to be acid-resistant, the presence of some acid-resistant strains in silage, human digestive tract and other extremely acidic environments has still been reported [40,41,42,43]. Previous studies have shown that the E. coli O157:H7 strain may produce acid resistance by inducing acid-tolerant response, so it can survive in an environment with a pH as low as 3.4 [44, 45]. LS especially WA inoculation further reduced the potential risk of Staphylococcus aureus infection, pathogenic E. coli infection and shigellosis, which could be attributed to their aforementioned antibacterial effects. The production of antimicrobial metabolites such as acetic acid may account for their antibacterial effects. While, it is interesting to note that WA has lower acetic acid concentration but higher antibacterial performance than LS. We speculated that the secretion of bioactive compounds such as bacteriocin and extracellular polysaccharides could be the main reason for the antibacterial function of WA [46,47,48,49], so further research is needed thereafter. The pathogenic modules showed similar characteristics to the pathogenic pathways, that is, WA showed the best performance in reducing the pathogenicity signature of EHEC/EPEC and Bacillus anthracis.

The microbiota of fresh herbage or silage is increasingly being linked to acute and chronic diseases in animals [38, 50]. The foodborne pathogen influences various aspects of animal health, including immune response, metabolism, disease occurrence, and finally body status and behavior [51, 52]. Therefore, it's important to clarify the microbial phenotypes in the microbiome, especially their pathogenic potential [18]. For instance, identifying the ratio of Gram stain can help develop more targeted antibiotic treatment approaches [53]. Furthermore, determining the anaerobic bacteria and pathogen levels in feedstuff can, at sources, prevent potential pathogens from spreading on farms or from animal products to humans. Nevertheless, numerous current tools, including popular PICRUSt, are unable to deduce prevalent microbial phenotypes from sequencing data, such as ‘gram negative’, ‘gram positive’, ‘forms biofilms’, ‘contains mobile elements’, ‘stress tolerant’, and ‘potentially pathogenic’. To address this, the BugBase tool was adopted to predict bacterial phenotypes relevant to the fresh herbage and silage. BugBase has already been applied to various microbiome datasets, including in precision medicine, agriculture, food and environmental research [18]. It has proven useful in identifying interpretable microbiological discoveries that current tools cannot achieve. When combined with 16S rRNA gene sequencing data, BugBase can provide new insights into the pathogenic risks and safety of silage microbiota, making it a valuable tool for assessing the safety of silages.

In this study, the ensiling process effectively removed the bacterial communities that were mobile-elements-containing and potentially pathogenic, which was due to the effective anaerobic maintenance and fast acidification suppressed a large number of harmful microbes (such as Staphylococcus, Mammaliicoccus and Brucella). The higher relative abundance of gram-positive in LS and WA could be attributed to the promotion of LABs after inoculation since LABs are gram-positive bacteria [54, 55]. After 30 days of ensiling, the negligible relative abundance of bacterial communities that are biofilm-formed, potentially pathogenic and stress-tolerant after LS and WA inoculation showed that, at the bacterial phenotypic level, the inoculation of LS and WA is also able to eliminate pathogenic concerns of WCC silage. The decrease in the relative abundance of gram-negative and biofilm formation accompanied by a decrease in the relative abundance of potential pathogenic is desirable for WCC silage production because gram-negative bacteria with biofilms such as lipopolysaccharide layers are tricky and have been reported to cause serious systemic infections [56, 57]. Ensilage did decrease the proportion and risk of pathogenic bacteria, and pathogens and these negative effects could be further dispelled by LS and WA inoculation.

Given that these pathogenic predictions are based on species assignments of a low-resolution target (97% OTU clustering method), especially pathogenic Staphylococcus aureus and nonpathogenic Staphylococcus epidermidis share > 98.5% nucleotide sequence identity, the risk of Staphylococcus aureus infection are highly speculative and should be examined with caution. The predicted pathogenic pathways, such as abovementioned Staphylococcus aureus infection, pathogenic E. coli infection and shigellosis, can be further validated by high-resolution approach (e.g. Amplicon Sequence Variant).

Conclusions

Ensiling process and LAB inoculation both had effectively improved the quality of WCC silage in terms of fermentation characteristics, bacterial diversity, co-occurrence networks, predicted functionality and phenotype. Short-term ensilage did reduce the proportion of pathogenic bacteria in WCC material, but was not effective for inhibiting some acid-resistant pathogens, while these risks could be eliminated by inoculation with W. confuse. Therefore, W. confuse could be developed as a promising fast start-up inoculant for reducing pathogenic contamination and improving hygienic quality of silage.

Availability of data and materials

The raw sequenced data were deposited in the NCBI Sequence Read Archive (SRA) database under the accession no. PRJNA1073498.

Abbreviations

WCC:

Whole-crop corn

LAB:

Lactic acid bacteria

KEGG:

Kyoto Encyclopedia of Genes and Genomes

PCR:

Polymerase chain reaction

PCoA:

Principal coordinate analysis

GLM:

General linear model

ANOVA:

Analysis of variance

LA/AA:

Lactic acid/acetic acid

OTUs:

Operational Taxonomic Units

FM:

Fresh WCC

CON:

Control (WCC treated with distilled water)

LS:

WCC inoculated with Lc. Lactis

WA:

WCC inoculated with W. confuse

EHEC/EPEC:

Enterohemorrhagic E. coli/enteropathogenic E. coli

References

  1. Dong J, Li S, Chen X, Sun Z, Sun Y, Zhen Y, Qin G, Wang T, Demelash N, Zhang X. Effects of Lactiplantibacillus plantarum inoculation on the quality and bacterial community of whole-crop corn silage at different harvest stages. Chem Biol Technol Agric. 2022;9:57.

    Article  CAS  Google Scholar 

  2. Bai J, Franco M, Ding Z, Hao L, Ke W, Wang M, Xie D, Li Z, Zhang Y, Ai L, Guo X. Effect of Bacillus amyloliquefaciens and Bacillus subtilis on fermentation, dynamics of bacterial community and their functional shifts of whole-plant corn silage. J Anim Sci Biotechnol. 2022;13:7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  3. Dong Z, Li J, Chen L, Wang S, Shao T. Effects of freeze-thaw event on microbial community dynamics during red clover ensiling. Front Microbiol. 2019;10:1559.

    Article  PubMed  PubMed Central  Google Scholar 

  4. Zhao J, Yin X, Wang S, Li J, Shao T. Separating the effects of chemical and microbial factors on fermentation quality and bacterial community of Napier grass silage by using gamma-ray irradiation and epiphytic microbiota transplantation. Anim Feed Sci Technol. 2021;280: 115082.

    Article  CAS  Google Scholar 

  5. McDonald P, Henderson AR, Heron SJE. The biochemistry of silage. 2nd ed. Bucks: Chalcombe Publications; 1991.

    Google Scholar 

  6. Ferrero F, Piano S, Tabacco E, Borreani G. Effects of conservation period and Lactobacillus hilgardii inoculum on the fermentation profile and aerobic stability of whole corn and sorghum silages. J Sci Food Agric. 2019;99:2530–40.

    Article  CAS  PubMed  Google Scholar 

  7. Kleinschmit DH, Kung L Jr. The effects of Lactobacillus buchneri 40788 and Pediococcus pentosaceus R1094 on the fermentation of corn silage during various stages of ensiling. J Dairy Sci. 2006;89:3999–4004.

    Article  CAS  PubMed  Google Scholar 

  8. Kung L, Smith ML, Benjamim Silva E, Windle MC, da Silva TC, Polukis SA. An evaluation of the effectiveness of a chemical additive based on sodium benzoate, potassium sorbate, and sodium nitrite on the fermentation and aerobic stability of corn silage. J Dairy Sci. 2018;101:5949–60.

    Article  CAS  PubMed  Google Scholar 

  9. Boyaci Gunduz CP, Gaglio R, Franciosi E, Settanni L, Erten H. Molecular analysis of the dominant lactic acid bacteria of chickpea liquid starters and doughs and propagation of chickpea sourdoughs with selected Weissella confusa. Food Microbiol. 2020;91: 103490.

    Article  CAS  PubMed  Google Scholar 

  10. Tenea GN, Lara MI. Antimicrobial compounds produced by Weissella confusa Cys2-2 strain inhibit Gram-negative bacteria growth. CyTA-J Food. 2019;17:105–11.

    Article  CAS  Google Scholar 

  11. Fang X, Li Y, Guo W, Ke W, Bi S, Guo X, Zhang Y. Lactobacillus delbrueckii subsp. bulgaricus F17 and Leuconostoc lactis H52 supernatants delay the decay of strawberry fruits: a microbiome perspective. Food Funct. 2019;10:7767–81.

    Article  CAS  PubMed  Google Scholar 

  12. Hwang I-C, Oh JK, Kim SH, Oh S, Kang D-K. Isolation and characterization of an anti-listerial bacteriocin from Leuconostoc lactis SD501. Korean J Food Sci An. 2018;38:1008–18.

    Article  Google Scholar 

  13. Lakra AK, Domdi L, Hanjon G, Tilwani YM, Arul V. Some probiotic potential of Weissella confusa MD1 and Weissella cibaria MD2 isolated from fermented batter. Lwt. 2020;125: 109261.

    Article  CAS  Google Scholar 

  14. Lee S, Park J, Jang J-K, Lee B-H, Park Y-S. Structural analysis of gluco-oligosaccharides produced by Leuconostoc lactis and their prebiotic effect. Molecules. 2019;24:3998.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Dey DK, Koo BG, Sharma C, Kang SC. Characterization of Weissella confusa DD_A7 isolated from kimchi. Lwt. 2019;111:663–72.

    Article  CAS  Google Scholar 

  16. Dey DK, Khan I, Kang SC. Anti-bacterial susceptibility profiling of Weissella confusa DD_A7 against the multidrug-resistant ESBL-positive E. coli. Microb Pathog. 2019;128:119–30.

    Article  CAS  PubMed  Google Scholar 

  17. McAllister TA, Duniere L, Drouin P, Xu S, Wang Y, Munns K, Zaheer R. Silage review: using molecular approaches to define the microbial ecology of silage. J Dairy Sci. 2018;101:4060–74.

    Article  CAS  PubMed  Google Scholar 

  18. Ward T, Larson J, Meulemans J, Hillmann B, Lynch J, Sidiropoulos D, Spear JR, Caporaso G, Blekhman R, Knight R, Fink R, Knights D. BugBase predicts organism-level microbiome phenotypes. bioRxiv. 2017. https://doi.org/10.1101/133462.

    Article  Google Scholar 

  19. Broderick GA, Kang JH. Automated simultaneous determination of ammonia and total amino acids in ruminal fluid and in vitro media. J Dairy Sci. 1980;63:64–75.

    Article  CAS  PubMed  Google Scholar 

  20. Hagberg AA, Schult DA, Swart PJ. Exploring network structure, dynamics, and function using NetworkX. In: Varoquaux G, Vaught T, Millman J, editors. Proceedings of the 7th Python in Science conference. Pasadena: SciPy; 2008. p. 11–5.

    Google Scholar 

  21. Gloor GB, Macklaim JM, Pawlowsky-Glahn V, Egozcue JJ. Microbiome datasets are compositional: and this is not optional. Front Microbiol. 2017;8:2224.

    Article  PubMed  PubMed Central  Google Scholar 

  22. Bastian M, Heymann S, Jacomy M. Gephi: an open source software for exploring and manipulating networks. Proc Int AAAI Conf Web Soc Media. 2009;3:361–2.

    Article  Google Scholar 

  23. Berry D, Widder S. Deciphering microbial interactions and detecting keystone species with co-occurrence networks. Front Microbiol. 2014;5:219.

    Article  PubMed  PubMed Central  Google Scholar 

  24. Douglas GM, Maffei VJ, Zaneveld JR, Yurgel SN, Brown JR, Taylor CM, Huttenhower C, Langille MGI. PICRUSt2 for prediction of metagenome functions. Nat Biotechnol. 2020;38:685–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  25. Neylon JM, Kung L Jr. Effects of cutting height and maturity on the nutritive value of corn silage for lactating cows. J Dairy Sci. 2003;86:2163–9.

    Article  CAS  PubMed  Google Scholar 

  26. Cai Y, Benno Y, Ogawa M, Ohmomo S, Kumai S, Nakase T. Influence of Lactobacillus spp. from an inoculant and of Weissella and Leuconostoc spp. from forage crops on silage fermentation. Appl Environ Microbiol. 1998;64:2982–7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. Guo XS, Ke WC, Ding WR, Ding LM, Xu DM, Wang WW, Zhang P, Yang FY. Profiling of metabolome and bacterial community dynamics in ensiled Medicago sativa inoculated without or with Lactobacillus plantarum or Lactobacillus buchneri. Sci Rep. 2018;8:357.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  28. Kamphayae S, Kumagai H, Angthong W, Narmseelee R, Bureenok S. Effects of different ratios and storage periods of liquid brewer’s yeast mixed with cassava pulp on chemical composition, fermentation quality and in vitro ruminal fermentation. Asian Australas J Anim Sci. 2017;30:470–8.

    Article  CAS  PubMed  Google Scholar 

  29. Ogunade IM, Jiang Y, Pech Cervantes AA, Kim DH, Oliveira AS, Vyas D, Weinberg ZG, Jeong KC, Adesogan AT. Bacterial diversity and composition of alfalfa silage as analyzed by Illumina MiSeq sequencing: effects of Escherichia coli O157:H7 and silage additives. J Dairy Sci. 2018;101:2048–59.

    Article  CAS  PubMed  Google Scholar 

  30. Kung L Jr, Shaver RD, Grant RJ, Schmidt RJ. Silage review: Interpretation of chemical, microbial, and organoleptic components of silages. J Dairy Sci. 2018;101:4020–33.

    Article  CAS  PubMed  Google Scholar 

  31. Jing Y, Mu C, Wang H, Shen J, Zoetendal EG, Zhu W. Amino acid utilization allows intestinal dominance of Lactobacillus amylovorus. ISME J. 2022;16:2491–502.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  32. Zadoks RN, Griffiths HM, Munoz MA, Ahlstrom C, Bennett GJ, Thomas E, Schukken YH. Sources of Klebsiella and Raoultella species on dairy farms: be careful where you walk. J Dairy Sci. 2011;94:1045–51.

    Article  CAS  PubMed  Google Scholar 

  33. Zou X, Zhou W, Wu S, Chen X, Zhang Q. Dynamic changes in the bacterial communities and metabolites of Moringa oleifera leaves during fermentation with or without pyroligneous acid. Lwt. 2023;177: 114593.

    Article  CAS  Google Scholar 

  34. Banerjee S, Kirkby CA, Schmutter D, Bissett A, Kirkegaard JA, Richardson AE. Network analysis reveals functional redundancy and keystone taxa amongst bacterial and fungal communities during organic matter decomposition in an arable soil. Soil Biol Biochem. 2016;97:188–98.

    Article  CAS  Google Scholar 

  35. Bolsen KK, Lin C, Brent BE, Feyerherm AM, Urban JE, Aimutis WR. Effect of silage additives on the microbial succession and fermentation process of alfalfa and corn silages. J Dairy Sci. 1992;75:3066–83.

    Article  CAS  Google Scholar 

  36. Benjamim da Silva É, Liu X, Mellinger C, Gressley TF, Stypinski JD, Moyer NA, Kung L. Effect of dry matter content on the microbial community and on the effectiveness of a microbial inoculant to improve the aerobic stability of corn silage. J Dairy Sci. 2022;105:5024–43.

    Article  CAS  PubMed  Google Scholar 

  37. Banerjee S, Schlaeppi K, van der Heijden MGA. Keystone taxa as drivers of microbiome structure and functioning. Nat Rev Microbiol. 2018;16:567–76.

    Article  CAS  PubMed  Google Scholar 

  38. Queiroz OCM, Ogunade IM, Weinberg Z, Adesogan AT. Silage review: foodborne pathogens in silage and their mitigation by silage additives. J Dairy Sci. 2018;101:4132–42.

    Article  CAS  PubMed  Google Scholar 

  39. Zhang SL, Maddox CW. Cytotoxic activity of coagulase-negative staphylococci in bovine mastitis. Infect Immun. 2000;68:1102–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Chen Y, Sela S, Gamburg M, Pinto R, Weinberg ZG. Fate of Escherichia coli during ensiling of wheat and corn. Appl Environ Microbiol. 2005;71:5163–70.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Ogunade IM, Kim DH, Jiang Y, Weinberg ZG, Jeong KC, Adesogan AT. Control of Escherichia coli O157:H7 in contaminated alfalfa silage: effects of silage additives. J Dairy Sci. 2016;99:4427–36.

    Article  CAS  PubMed  Google Scholar 

  42. Arnold KW, Kaspar CW. Starvation-induced and stationary-phase-induced acid tolerance in Escherichia-Coli O157/H7. Appl Environ Microbiol. 1995;61:2037–9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  43. Buchanan RL, Edelson SG. pH-dependent stationary-phase acid resistance response of enterohemorrhagic Escherichia coli in the presence of various acidulants. J Food Prot. 1999;62:211–8.

    Article  CAS  PubMed  Google Scholar 

  44. Brudzinski L, Harrison MA. Influence of incubation conditions on survival and acid tolerance response of Escherichia coli O157:H7 and non-O157:H7 isolates exposed to acetic acid. J Food Prot. 1998;61:542–6.

    Article  CAS  PubMed  Google Scholar 

  45. Koutsoumanis KP, Sofos JN. Comparative acid stress response of Listeria monocytogenes, Escherichia coli O157:H7 and Salmonella Typhimurium after habituation at different pH conditions. Lett Appl Microbiol. 2004;38:321–6.

    Article  CAS  PubMed  Google Scholar 

  46. Adesulu-Dahunsi AT, Sanni AI, Jeyaram K, Ojediran JO, Ogunsakin AO, Banwo K. Extracellular polysaccharide from Weissella confusa OF126: production, optimization, and characterization. Int J Biol Macromol. 2018;111:514–25.

    Article  CAS  PubMed  Google Scholar 

  47. Malik A, Yuliantie E, Suprahman NY, Linardi T, Widiyanti AW, Haldy J, Tjia C, Takagi H. Construction and functional analysis of the recombinant bacteriocins Weissellicin-MBF from Weissella confusa MBF8-1. Curr Pharm Biotechnol. 2021;22:115–22.

    Article  CAS  PubMed  Google Scholar 

  48. Malik A, Sumayyah S, Yeh C-W, Heng NCK, Jass J. Identification and sequence analysis of pWcMBF8–1, a bacteriocin-encoding plasmid from the lactic acid bacterium Weissella confusa. FEMS Microbiol Lett. 2016;363:fnw059.

    Article  PubMed  Google Scholar 

  49. Juvonen R, Honkapää K, Maina NH, Shi Q, Viljanen K, Maaheimo H, Virkki L, Tenkanen M, Lantto R. The impact of fermentation with exopolysaccharide producing lactic acid bacteria on rheological, chemical and sensory properties of pureed carrots (Daucus carota L.). Int J Food Microbiol. 2015;207:109–18.

    Article  CAS  PubMed  Google Scholar 

  50. Driehuis F, Wilkinson JM, Jiang Y, Ogunade I, Adesogan AT. Silage review: animal and human health risks from silage. J Dairy Sci. 2018;101:4093–110.

    Article  CAS  PubMed  Google Scholar 

  51. Allard SM, Micallef SA. The plant microbiome: diversity, dynamics, and role in food safety. London: Academic press; 2019. p. 229–57.

    Google Scholar 

  52. Okoye CO, Wang Y, Gao L, Wu Y, Li X, Sun J, Jiang J. The performance of lactic acid bacteria in silage production: a review of modern biotechnology for silage improvement. Microbiol Res. 2023;266: 127212.

    Article  CAS  PubMed  Google Scholar 

  53. Langdon A, Crook N, Dantas G. The effects of antibiotics on the microbiome throughout development and alternative approaches for therapeutic modulation. Genome Med. 2016;8:39.

    Article  PubMed  PubMed Central  Google Scholar 

  54. Axelsson L. Lactic acid bacteria: classification and physiology. In: Salminen S, Von Wright A, editors. Lactic acid bacteria. New York: Marcel Dekker; 2004. p. 1–63.

    Google Scholar 

  55. Li J, Ma D, Tian J, Sun T, Meng Q, Li J, Shan A. The responses of organic acid production and microbial community to different carbon source additions during the anaerobic fermentation of Chinese cabbage waste. Bioresour Technol. 2023;371: 128624.

    Article  CAS  PubMed  Google Scholar 

  56. Wang Z, Li Q, Li J, Shang L, Li J, Chou S, Lyu Y, Shan A. pH-responsive antimicrobial peptide with selective killing activity for bacterial abscess therapy. J Med Chem. 2022;65:5355–73.

    Article  CAS  PubMed  Google Scholar 

  57. Zhao J, Jing ZD, Yin XJ, Wang SR, Li JF, Dong ZH, Shao T. Zero-cost decision-making of mowing height selection for promoting cleaner and safer production in the feed industry. J Clean Prod. 2023;429:139451. https://doi.org/10.1016/j.jclepro.2023.139451.

    Article  CAS  Google Scholar 

Download references

Acknowledgements

Not applicable.

Funding

This work was financially supported by the NSFC (32201464), the Fundamental Research Funds for the Central Universities (KYQN2023001) and the Jiangsu Funding Program for Excellent Postdoctoral Talent (2022ZB319).

Author information

Authors and Affiliations

  1. Institute of Ensiling and Processing of Grass, College of Agro-Grassland Science, Nanjing Agricultural University, Nanjing, 210095, People’s Republic of China

    Jie Zhao, Xinbao Li, Haopeng Liu, Zhaodi Jing, Xuejing Yin, Junfeng Li, Zhihao Dong & Tao Shao

Authors
  1. Jie Zhao
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Xinbao Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Haopeng Liu
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Zhaodi Jing
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Xuejing Yin
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Junfeng Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Zhihao Dong
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Tao Shao
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

TS and YJ designed the study, contributed to securing financial support. JZ and ZJ conducted experiment. JZ prepared the manuscript; JL, ZD, and TS revised the manuscript draft; XY performed data collection and statistical analysis. All authors have read and approved the final manuscript.

Corresponding author

Correspondence to Tao Shao.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Zhao, J., Li, X., Liu, H. et al. Ensilage using Leuconostoc lactis and Weissella confusa reduces microbial risk and enhances hygienic quality of whole-crop corn. Chem. Biol. Technol. Agric. 11, 44 (2024). https://doi.org/10.1186/s40538-024-00566-0

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s40538-024-00566-0

Keywords