Effects of lactic acid bacteria and molasses additives on the microbial community and fermentation quality of soybean silage

Accepted Manuscript Effects of lactic acid bacteria and molasses additives on the microbial community and fermentation quality of soybean silage Kuikui Ni, Fangfang Wang, Baoge Zhu, Junxiang Yang, Guoan Zhou, Yi Pan, Jin Zhong PII: DOI: Reference:

S0960-8524(17)30545-X http://dx.doi.org/10.1016/j.biortech.2017.04.055 BITE 17950

To appear in:

Bioresource Technology

Received Date: Revised Date: Accepted Date:

23 February 2017 13 April 2017 15 April 2017

Please cite this article as: Ni, K., Wang, F., Zhu, B., Yang, J., Zhou, G., Pan, Y., Zhong, J., Effects of lactic acid bacteria and molasses additives on the microbial community and fermentation quality of soybean silage, Bioresource Technology (2017), doi: http://dx.doi.org/10.1016/j.biortech.2017.04.055

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Effects of lactic acid bacteria and molasses additives on the microbial community and fermentation quality of soybean silage

Kuikui Nia, Fangfang Wanga,c, Baoge Zhub, Junxiang Yangd, Guoan Zhou b, Yi Panb, Jin Zhonga,c* a

State Key Laboratory of Microbial Resources, Institute of Microbiology, Chinese Academy of Sciences, Beijing 100101, China

b

State Key Laboratory of Plant Cell and Chromosome Engineering, Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, Beijing 100101, China

c

School of Life Science, University of Chinese Academy of Sciences, Beijing 100039, China

d

National Animal Husbandry Station, Beijing 100022, China

*Corresponding author. E-mail address: [email protected] (J. Zhong).

HIGHLIGHTS






Soybean ensiled with lactic acid bacteria inoculant and molasses. All additives improved fermentation quality of soybean silage. Combined addition of lactic acid bacteria and molasses showed best silage quality. Molasses enriched the abundance of Lactobacillus. Combination of lactic acid bacteria and molasses decreased Clostridia abundance.

ABSTRACT The objective was to study effects of lactic acid bacteria (L) and molasses (M) on the microbial community and fermentation quality of soybean silage. Soybean was ensiled with no additive control (C), 0.5% molasses (0.5%M), 0.5%M+L (0.5%ML), 2%M, 2%M+L (2%ML) for 7, 14, 30 and 60 days. The M-treated silages could increase the content of lactic acid and decrease butyric acid than control. Besides, higher crude protein was also observed in M-treated silages. With prolonged ensiling time, there was a reduction of the ratio of lactic acid/acetic acid in the 2%M-treated and 2%ML-treated silages. The combined addition of L and 2%M could enhance the account of desirable Lactobacillus and inhibit the growth of undesirable microorganism such as Clostridia and Enterobacter. In summary, the silage quality of soybean was improved with the addition of L and M. Keywords: Soybean; Silage; Lactic acid bacteria; Molasses

1. Introduction With the rapid development of economy in China, the demand for animal products is increasing greatly. The main constraint factor for further improvement of animal products is the lack of adequate and high quality green fodder for animal feed. Soybean (Glycine max Merr.) with relatively high content of protein has been considered to be a promising source of protein, instead of costly method of supplying feeding diet with protein concentration (Jahanzad et al., 2016). However, soybean harvest is seasonal with high accumulation, which need safe and efficient conservation way for ruminants. Ensiling is a common way to preserve the moist forage crop, which could prolong the storage time and improve the feed palatability via lactic acid fermentation under anaerobic condition. During ensiling process, lactic acid bacteria and water soluble carbohydrate (WSC) are crucial factors for high silage quality (McDonald et al., 1991; Cai et al., 1998). Among the various additives, L inoculant and M have been proposed as an effective stimulant to improve silage quality through increasing the initial L load and fermentable substrate, respectively (Li et al., 2014; Ni et al., 2015). Besides, until now few publications are related to their application on soybean silage. The use of soybean silage as animal feed is advantageous, such as rich protein in soybean reducing the dependence on the fluctuating price of protein in the market. Moreover, soybean has a strong growth ability to be tolerance of different environment. However, natural fermentation of soybean silage usually lead to unpleasant odour and high butyric acid content, possible due to low level of WSC content (Budakli et al., 2016). Unlike alfalfa or whole crop

corn silages, little effort is devoted to improving the silage quality of soybean. Recently, some researchers found silage produced from the mixture of millet and soybean exhibited enhanced fermentation, indicated by lower pH and higher lactic acid than silage from sole soybean (Jahanzad et al., 2016). However, little information is related to the microbial community and fermentation products during ensiling process of soybean, which might provide important information for further regulation of fermentation. Therefore, the present study was to investigate the effects of L inoculant and M on microbial community and fermentation characteristics during the ensiling process of soybean.

2. Materials and methods

2.1. Forage harvest and silage preparation

Soybean was cultivated and harvested on 25 August 2016 (sample 1: podding stage) and on 10 September 2016 (sample 2: early fruiting stage) from experimental field of Chinese Academy of Sciences (116°24′E, 40°06′ N). The crop was chopped to the theoretical length of 2 cm using a crop chopper. L inoculant and molasses were used as additives for ensiling soybean. L inoculant was a combination of Lactobacillus plantarum and Pediococcus pentosaceus (applied at a ratio of 4:1), and both were isolated form corn silage. The combined inoculant was added at a level of 106 colony forming unit (cfu) per gram of fresh material (FM). The chopped soybean was mixed and divided into equal portion for six treatments: No additive

control (C), L treatment (L), 0.5g Molasses per 100g FM (0.5%M), 0.5%M+L (0.5%ML), 2g Molasses per 100 FM (2%M) and 2% M+L (2%ML). All the employed additives were mixed homogenous with soybean, then 2.5±0.1kg chopped soybean was packed and compressed manually into a 1.5L jar. A total of 144 jars (2 materials×4 ensiling days×6 treatments×3 repeat) were made and kept at ambient temperature (21-30oC)。 Three jars for each treatment were opened for analyzing microbial community and organic acid after 7, 14, 30 and 60 days of ensiling, respectively.

2.2. Analysis of microbial population, organic acid and chemical composition

The jars were opened in a clean bench, the samples (10 g) were blended with 90ml of sterilized water, and serially diluted from 10-1 to 10-5 in sterilized water. The number of lactic acid bacteria were measured by plate count on lactobacilli de Man, Rogosa, Sharpe (MRS) agar incubated at 30oC for 48 h under anaerobic conditions (DG 250/min MACS; Don Whitley Science; England). Molds and yeasts were counted on potato dextrose agar (Nissui), incubated at 30oC for 24 h, and yeasts were distinguished from molds and other bacteria by colony appearance and the observation of cell morphology. Coliform was counted on nutrient agar (Nissui), incubated at 30oC for 24 h under aerobic conditions. Colonies were counted as viable numbers of microorganisms in cfu/g FM-1. For determination of pH and organic acid, 3 g fresh samples with 30 mL sterilized water was homogenized in a blender for 1 minute, and then filtrated through 0.22µm membrane filters. The

pH of this extract was immediately measured with a glass electrode pH meter (pH 213; HANNA; Italy). Concentrations of organic acid were measured in HPLC (1200, Agilent, America) fitted with a UV detector (210 nm) and a column (ICSep COREGEL-87H). The mobile phase was 0.005 M H2SO4 at a flow rate of 0.6 mL/min at 55°C. Principal component analysis (PCA) based on pH values, microbial population and the content of organic acid during ensiling process was performed using the JMP software (ver. 10; SAS Institute, Tokyo, Japan). DM contents were determined by oven drying at 60°C for 48 h. The samples were dried in a forced-air oven at 60oC for 48 h and ground to pass a 1-mm screen with a Wiley mill (ZM200, Retsch GmbH). Crude protein and ether extract (EE) were analyzed according to standard procedures detailed by the Association of Official Analytical Chemists (AOAC 1990). The neutral detergent fiber (NDF) and acid detergent fiber (ADF) were analyzed by the method of Van Soest et al. (1991).

2.3. Microbial diversity analysis

2.3.1.

Microbial DNA isolation

For the molecular analysis of the microbial communities, genomic DNA was extracted from following steps: 50 g silage mixed with 200 mL sterile 0.85% NaCl solution, then treated with table concentrator at 120 r/m for 2 h. Filtered with carbasus and the centrifuged the liquor at 10,000 r/m for 10 min at 4oC, discarded the supernatant and suspended the deposit in 3 mL

sterile 0.85% NaCl solution. The liquor was then centrifuged at 10,000 r/m for 10 min at 4oC, and the supernatant was discarded. The microbial DNAs were isolated from the deposit using a Fast DNATM SPN Kit. The aliquoted DNA samples were kept at-20o C until further analysis.

2.3.2. PCR amplification The DNA samples used in the high-throughput sequencing were amplified using primers targeting the V3-V4 regions of 16S rRNA genes (forward: 5′ACTCCTACGGGAGGC AGCAG-3′; reverse:5′GGACTACHVGGGTWTCTAAT-3′). Approximately 30ng of DNA isolated from each sample was used for amplification. Polymerase chain reactions were carried out under the following condition: a prior denaturation at 95oC for 25 min, followed by 25 cycles of denaturation at 95oC for 30 s, annealing at 56oC for 30s, elongation at 72oCfor 40 s and extension at 72o C for 10 min. The reaction was performed in a 50 mixture containing 5 µL of 10×Pyrobest Buffer, 4 µL of 2.5 mM dNTPs, 2 µL of each primer (10µM), 0.3 µL Pyrobest DNA Polymerase (2.5 U/µL, TaKaRa Code: DR005A), and 30 ng of template DNA. To reduce PCR deviation, PCR reaction for one treatment was performed in triplicate. Then, their mixture was used to perform sequencing.

2.3.3. High-throughput sequencing of metagenomic DNA The DNA samples were sequenced at the Auwigene Company using Paired-end sequencing with an Illumina MiSeq PE300 platform. In order to get high quality sequencing, their barcode and primers were discharged. And then Mothur (v.1.34.4) was used to discharge the sequences

less than 200bp whose maxhomop value is greater than 10. Remaining sequences were checked for chimeras in the de novo mode by USEARCH 8.0 (Edgar, 2010). After filtering process, the clean-tag remained for downstream analysis. The operational taxonomic units (OTUs) at 97% similarity level were clustered using QIIME (v1.8.0). OTUs file was used to calculate rarefaction (R (v.22)) and alpha diversity (Mothur (v1.34.4)).

2.4. Statistical analyses

The statistical analyses were performed using JMP software (version 10; SAS Institute, Tokyo, Japan) to examine the differences between different treatment. Tukey,s HSD test was employed to the differences the treatment means and the significance was declared at P<0.05.

3. Results and discussion

3.1. The chemical composition and microbial population by plate culture before ensiling

As shown in Table 1, the DM contents of sample 1 and sample 2 were 24.86% and 27.22%, respectively. Liu et al. (2016) reported that later stage had lower moisture content than that harvested at an earlier stage, thus soybean harvested at later stage (sample 2) may reduce the cost of biomass transportation. The CP concentration was around 14% and the EE concentration was 1.2-1.5%, which was lower compared with the data (CP:17%) by Jahanzad et al. (2014).

Several factors can influence forage nutrition, such as plant genotype, sowing density, harvest season, irrigation and fertilization (Vasco-Correa et al., 2015). The relatively low CP could be explained by the poor fertilization in this study. Inclusion with soybean can reduce the NDF content in mixture silage, probably because soybean is less fibrous (Paterson et al., 1994; Sibanda et al., 1997). WSC was an important factor for silage fermentation, and the content greater than 5%DM was shown to be crucial for assuring acceptable fermentation quality. The WSC content in this study was as low as 1% in both growth stages, which indicated difficulty in good preservation of soybean without any addition through ensiling method. The population of the initial L, yeast, coliform and mold was similar between sample 1 and sample 2. The L number was 104 cfu/g FM-1 while the yeast and coliform number ranged around 106 cfu/g FM-1. Generally, lactic acid bacteria number is considered as a crucial factor in predicating the adequacy of silage fermentation and in determining whether to apply bacterial inoculant to silage materials. When L number reaches over 105 cfu/g FM-1, silage can be well preserved (Cai et al., 1998). The low L number and high harmful microorganism might indicate silage fermentation of soybean need to be controlled by L inoculation or extra sugar.

3.2. Effect of L and M on pH values and microbial population of soybean silages

The dynamics of pH and microbial population were shown in Fig.1 and Fig.2. Overall, the pH values and microbial population of sample 1 and sample 2 showed similar trend during ensiling process. The factorial analysis revealed that L, M and L×M had effects of pH values and

coliform population while all these did not have an effect on L population (Table 3). The rate of pH decline is considered as an important indicator for reflecting the microbial activity and silage fermentation. In the whole ensilage time, all the treatments showed lower pH values than C, and a combination of L and M led to the highest pH drop among the treatments. The drop of pH values mainly occurred in the first 7 or 14 days of ensiling, whereas no furthrer significant pH reduction was observed with prolonged ensilage time, which was in accordance with the report by Desta et al. (2016). The final pH values in 2%M and 2%ML treatments declined to below 4.5, which could be the direct increasement of fermentable substrate promoting sufficeient lactic acid production and preserve silage by low pH. However, the pH value of C still maintained around 6.0. The above results further proved addition of L and M at ensiling process could stimulate the pH drop. In addition, sample 2 with L addition showed lower pH values than sample1 with L addition, possble due to the higher level of WSC in sample 2. Ensiling increased L number at the beginning, then the number decreased from 108 to 106 cfu/g FM-1. Similarly, Kim et al. (2016) found L population reduced as the ensiling duration of spent mushrom substrate increased. Unexpectively, the additions including L treatment did not exhibit higher level of L number compared with untreated group (Table 3). Although we could not clasify the behind reason clearly, the high number of the epiphytic microorganisms, time of ensiling and dose of bacterial applied are among factors that can affect the function of inoculants (Canibe et al., 2014). Ensiling also enhanced the yeast number at the beginning stage of ensiling, then decreased as the result of pH reduction. Overall, all the treatments showed

higher rate of yeast decline than C from the 7 day of ensiling, while the final yeast number still maitained at the level of 105 cfu/g FM-1. As we known, yeast could cause aerobic deterioration during feed-out stage and reduce the nutritional value of silage. Ni et al. (2015) repoted that the yeast number in whole crop rice silage ranged from 104-105 cfu/g FM-1 even though pH value dropped to below 4.2. Therefore, these results indicated further research into the factors controlling fungal growth in soybean silage is recommended. Coliform was found in the C and L treatment group during the whole ensiling process, but decreased to below the detectable level (102 cfu/g FM-1) after 7 or 14 days ensiling in other treatments.

3.3. Effect of L and M on the content of organic acid of soybean silages

As shown in Fig.3 and Fig.4, lactic acid and acetic acid were the dominant fermentation products in soybean silage, but propionic acid was not detected in all silages. L inoculant had effects on lactic acid and acetic acid contents, but not on butyric acid (Table 3). M and the interaction between L and M had highly effects on the contents of both lactic acid and butyric acid. Lactic acid is preferred over the ensiling process because it is the main organic acid responsible for pH reduction due to its strong acidity with a pKa 3.86. In this study, the overall lactic acid concentration increased intensively during the first 7 days of ensiling. In addition, lactic acid concentration began to decrease from 14 days in soybean silages treated with 2%M and 2%ML. The decrease in lactic acid and increase in acetic acid suggested that lactic acid was

converted into acetic acid with prolonged ensilage time, which was in accordance with the report by Alli et al. (1984) and Jahanzad et al. (2016). However, Desta et al. (2016) reported the plateau for Napier grass silage added with M happened at day 60. That discrepancy was probably due to Napier grass ensiled with more amount of M (4%FM) than our study, resulting in more soluble carbohydrate supplied for lactic acid bacteria metabolism. The ratio of lactic acid/acetic acid with 2%M and 2%ML were significantly higher than other treatments and C. The highest lactic acid concentrations were observed in soybean silages treated with 2%M and 2%ML silages while C and L-treated silages showed lowest lactic acid concentration. Compared with sole L treatment, M treatment significantly increased lactic acid concentration. This observation may indicate M addition contributed to L dominating the silage microbial community, subsequently directing the metabolism to a more a homo-fermentative pattern (Li et al., 2014). Butyric acid was produced by undesirable microorganism which break down amino acid resulting in nutrition loss. In this study, the contents of butyric acid in 2%M and 2%ML was lower than other treatments during the whole ensiling process, which was likely because their lower pH suppressed the growth of undesirable microorganism such as clostridia. The dynamic variance of microbial population and fermentation quality with different treatments can be demonstrated by principle component analysis. As shown in Fig.5, component 1 and component 2 can explained 51.3% and 24.2% of the total variance, respectively; Similarly, in Fig.6, component 1 and component 2 can explained 55.4% and 22.7%, respectively. The silage samples appeared to vary even more in untreated silages, depending on the sampling time. Basically, the 2%M and 2%ML treatments can be well separated from other treatment silages.

3.4. Effect of L and M on chemical composition of soybean silages after 60 days of ensiling

The chemical composition of soybean silages treated by different additives were shown in Table 2. Although different treatments had a significant effect on the contents of CP, EE and NDF, their contents varied in a very narrow range. Besides, all treatments did not have significant effects on the content of ADF. DM content was constantly higher in silage of sample 2. Higher contents of CP with 2%M and 2%ML could be attributed to the sharp pH reduction during the early stage of ensiling, then the growth of undesirable bacteria such as clostridia and the activity of plant enzyme were inhibited. In addition, silages treated with 2%ML showed lower NDF fraction, which could be explained by the acid hydrolysis of more digestible plan cell in the ensilage process (McDonald et al., 1991). All treatments in sample 1 silages showed lower NDF than C (Table 3), suggesting the relative immaturity materials were more susceptible to enzyme hydrolysis. Romero et al. (2016) found the similar phenomenon in oat silage.

3.5. Effect of L and M on bacterial community examined by NGS after 60 days of ensiling

NGS can provide a more detailed picture of bacterial community than conventional techniques (Ni et al., 2017). As far as we know this is the first report of the bacterial community in soybean silage.

As listed in Table 4, the recovered reads per sample ranged from 51569 to 69307. The coverage values of all samples were around 0.99, indicating most of bacteria were detected. A total of 2641 OTUs at 3% dissimilarity level were determined to further analyze the bacterial community. Chao, as another richness index of bacterial community, showed a similar trend with OTUs. The Shannon index of bacterial diversity was observed lowest in 2%ML-treated silages (1.7-1.9). Untreated soybean silages showed higher diversity compared with pre-ensiled material while greater diversity was found in the pre-ensiled crops by Ni et al. (2017). That different results were likely due to the relatively higher pH values in soybean silages could not suppress the microbial growth and subsequently reduce the bacterial diversity. The bacterial community of fresh soybean and silages at the genus level was shown in Fig.6. The dominant genus in the pre-ensiled crop were Enterobacter, Pantoea and Serraia, which were also detected in other materials such as guinea grass and alfalfa (Parvin et al., 2010; Li et al., 2011a; Nishino et al., 2012). However, their portions decreased greatly after ensiling, especially in the treated silages. Enterobacteria are non-spore forming, facultative anaerobe and could ferment lactic acid to acetic acid and other products, thus cause nutrition loss. The critical pH value for controlling the growth of Enterobacter in silage at 25%DM is 4.35 (Weissbach et al., 1996). In this study, Enterobacter abundance in untreated silage was still around 9%, which might account for its low silage quality. The presence of Serraia was usually associated with the production of 2,3-butanediol (McDonald et al., 1991), but we did not detect 2,3-butanediol in this study. Because even though Serraia appeared in the pre-ensiled soybean, it became marginal level after ensiling. Instead, Parvin et al. (2010) found Serraia was detected in the

ryegrass silage using denaturing gradient gel electrophoresis but not found in the pre-ensiled material, suggesting the distribution of Serraia was related to the type of material. The desirable lactic acid bacteria including Enterococcus (1.0-2.4%), Lactobacillus (0.5-0.7%), Pediococcus (0.1-0.2%) and Weissella (0.1%) was low in the pre-ensiled soybean. Bacillus was mainly resent in the pre-ensiled soybean, and its abundance decreased from 6.1% to 0.3% after 60 days of ensiling. Enterococcus was the dominant microbes in the C silages, followed by Pediococcus and Lactobacillus. Eikmeyer et al. (2013) found the Lactococcus, Leuconostoc and Weissella genera were the dominant in the untreated grass silage. Romero et al. (2017) reported Weissella, Leuconostoc and Pediococcus were the prevailing identifiable genera in the untreated oat silage. These results might indicate cocci-shaped lactic acid bacteria were usually present in the natural fermented silages. Lactobacillus and Pediococcus were the most abundance in the L-treated soybean silage, and the similar result was also reported by Bao et al. (2016), who ensiled Medicago sativa silage inoculated with Pediococcus acidilactici and Lactobacillus plantarum, and Pediococcus and Lactobacillus became the dominant genera when silage fermentation finished. M addition also contributed to the growth of Lactobacillus, especially for the combined addition of M and L, the Lactobacillus abundance reached to nearly 90% of the total population. As we known, lactic acid-producing cocci (Weissella, Leuconostocs, Pediococcus, Lactococci and Enterococci) initiate lactic fermentation at the early of ensiling process while lactic acid-rod (Lactobacillus) play an important role for pH reduction at the later stage (Cai et al., 1998). The high abundance of Lactobacillus in 2%ML silages could explain their relatively

well fermentation quality compared with other treatments (Li et al., 2011b; Jahanzad et al., 2016). Clostridia was often detected in silages, and its presence in silage is not desirable, because it can make use of protein and sugar to produce butyric acid resulting in bad silage quality with unpleasant palatability and low nutrition value. Therefore, inhibiting the growth of clostridia is important, especially in high nutrition silage such as soybean. In this study, the clostridia abundance in C and L silages ranged from 2.4 to 19.6%, while below 0.5% in silages treated with 2%ML, which indicated that the combined of M and L can suppress clostridia efficiently.

4. Conclusion

Additives could improve the silage quality of soybean at different extent, but soybean silages treated with 2%M and 2%ML had better fermentation quality than other treatments. The combined addition of M and L could enhance the abundance of desirable Lactobacillus and reduce the abundance of undesirable microorganism including Clostridia and Enterobacter. In summary, the combined addition of M and L could effectively improve the silage quality of soybean.

Acknowledgements

This work was supported by grants from Special Fund for Agro-scientific Research in the Public

Interest (201503134). Key research program of the Chinese Academy of Sciences (KFDZ-SW-101-4, KSZD-EW-Z-012-2, 3).

References

1. AOAC, 1990. Official Methods of Analysis. 15th ed. Association of Official Analytical Chemists, Arlington, VA. 2. Alli, I., Fairbairn, r., Norooz, I.E., Baker, B.B., 1984. The effects of molasses on the fermentation of chopped whole-plant leucaena. J. Sci. Food Sci. 35, 285-289. 3. Bao, W., Mi, Z., Xu, H., Zheng, Y., Kwok, L.Y., Zhang, H., Zhang, W.Y., 2016. Assessing quality of Medicago sativa silage by monitoring bacterial composition with single molecule, real-time sequencing technology and various physiological parameters. Sci. Rep. 6, 28358. 4. Budakli, C.E., 2016. Nutritive values of soybean silages ensiled with maize at different rates. Legume Res. 39, 810-813. 5. Cai, Y., Benno, Y., Ogawa, M., Ohmomo, S., Kumai, S., Nakase, K., 1998. Influence of Lactobacillus spp. from an inoculant and of Weissella and Leuconostoc spp. from forage crops on silage fermentation. Appl. Environ. Microb. 64, 2982-2987. 6. Canibe, N., Kristensen, N.B., Jensen, B.B.，Vils, E., 2014. Impact of silage additives on aerobic stability and characteristics of high-moisture maize during exposure to air, and on fermented liquid feed. J. Appl. Microbiol. 4, 747-760. 7. Desta, S.T., Yuan, X.J., Li, J.F., Shao, T., 2016. Ensiling characteristics, structural and

nonstructural carbohydrate composition and enzymic digestibility of Napier grass ensiled with additives. Bioresour. Technol. 221, 447-454. 8. Eikmeyer, F.G., Köfinger, P., Poschenel, A., Jünemann, S., Zakrzewski, M., Heinl, S., Mayrhuber, E., Grabherr, R., Pühler, A., Schwab, H., Schlüter, A., 2013. Metagenome analyses reveal the influence of the inoculant Lactobacillus buchneri CD034 on the microbial community involved in grass silaging. J. Biotechnol. 167, 334-343. 9. Jahanzad, E., Sadeghpour, A., Hashemi, M., Afshar, R., Hosseini, M.B., Barker, A., 2016. Silage fermentation profile, chemical composition and economic evaluation of millet and soya bean grown in monocultures and as intercrops. Grass Forage Sci. 71, 584-594. 10. Jahanzad, E., Sadeghpour, A., Hosseinim, B., Barker, A.V., Hashemi, M., Zandvakilio, R., 2014. Silage yield and nutritive value of millet-soybean intercrops as influenced by nitrogen application. Agron. J. 106, 1993-2000. 11. Kim, J.S., Lee, Y.H., Kim, Y.I., Ahmadi, F., Oh, Y.K., Park, J.M., Kwak, W.S., 2016. Effect of microbial inoculant or molasses in fermentative quality and aerobic stability of sawdust-based spent mushroom substrate. Bioresour. Technol. 216, 188-195. 12. Li, Y., Nishino, N., 2011a. Bacterial and fungal communities of wilted Italian ryegrass silage inoculated with and without Lactobacillus rhamnosus or Lactobacillus buchneri. Lett. Appl. Microbiol. 52, 314-321. 13. Li, Y., Nishino, N., 2011b. Monitoring the bacterial community of maize silage stored in a bunker silo inoculated with Enterococcus faecium, Lactobacillus plantarum and Lactobacillus buchneri. J. Appl. Microbiol. 110, 1561-1570.

14. Li, M., Zi, X.J., Zhou, H., Hou, G., Cai, Y., 2014. Effects of sucrose, glucose, molasses and cellulase on fermentation quality and in vitro gas production of king grass silage. Anim. Feed Sci. Technol. 197, 2016-212. 15. Liu, S., Ge, X., Liu, Z., Li, Y., 2016. Effect of harvest date on Arundo donax L. (giant reed) composition, ensilage performance, and enzymatic digestibility. Bioresour. Technol. 205, 97-103. 16. McDonald, P., Henderson, N., Herson, S., 1991. The Biochemistry of Silage. Chalcombe Publications, USA. 17. Ni, K., Wang, Y., Li, D., Cai, Y., Pang, H., 2015. Characterization, identification and application of lactic acid bacteria isolated from forage paddy rice silage. PLoS ONE. 10, e0121967. 18. Ni, K., Minh, T., Tsuruta, T., Pang, H., Nishino, N., 2017. Comparative microbiota assessment of wilted Italian ryegrass, whole crop corn, and wilted alfalfa silage using denaturing gradient gel electrophoresis and next-generation sequencing. Appl. Microbiol. Bio. 101, 1385-1394. 19. Nishino, N., Li, Y., Wang, C., Parvin, S., 2012. Effects of wilting and molasses addition on fermentation and bacterial community in guinea grass silage. Lett. Appl. Microbiol. 54, 175-181. 20. Paterson, J.A., Belyea, R.L., Bowman, J.P., Kerley, M.S., Williams, J.E., 1994. The impact of forage quality and supplementation regimen on ruminant animal intake and performance. In: Fahey G.C. Jr (ed). Forage quality, evaluation, and utilization, pp. 9-11. Madison, WI: ASA. 21. Parvin, S., Wang, C., Li, Y., Nishino, N., 2010. Effects of inoculation with lactic acid bacteria on the bacterial communities of Italian ryegrass, whole crop maize, guinea grass and rhodes grass silages. Anim. Feed Sci. Technol. 160, 160-166.

22. Romero, J.J., Zhao, Y., Balseca-Paredes, M.A., Tiezzi, F., Gutierrez-Rodriguez, E., Castillo, M.S., 2017. Laboratory silo type and inoculation effects on nutritional composition, fermentation, and bacterial and fungal communities of oat silage. J. Dairy Sci. 100, 1812-1828. 23. Sibanda. S., Jingura, R.M., Topps, J.H., 1997. The effect of level of inclusion of the legume Desmodium uncinatum and the use of molasses or ground maize as additives on the chemical composition of grass- and maize-legume silages. Anim. Feed Sci. Technol. 68, 295-305. 24. Van Soest, P.J., Robertson, J.B., Lewis, B.A., 1991. Methods for dietary fiber, neutral detergent fiber, and non-starch polysaccharides in relation to animal nutrition. J. Dairy Sci. 74, 3583-3597. 25. Vasco-Correa, J., Li, Y., 2015. Solid-state anaerobic digestion of fungal pretreated Miscanthus sinensis harvested in two different seasons. Bioresour. Technol. 185, 211-217. 26. Weissbach, F., Honig, H., 1996. Uber die Vorhersage und Steuerung des Garungsverlaufs bei der Silierung von Grunfutter aus extensivem Anbau. Landbauforsch. Volk. 1, 10-17.

Table 1 The chemical composition and microbial population by plate culture before ensiling. Soybean sample

DM

g kg-1 DM

Log cfu/g FM-1

CP ADF NDF EE WSC L Yeast Coliform Mold Sample 1 24.86 14.91 44.61 58.33 1.22 0.78 4.26 6.15 6.66 ND Sample 2 27.22 15.86 45.23 59.63 1.56 1.12 4.02 6.21 6.54 ND FM, fresh material; DM, dry matter; CP, crude protein; EE, ether extract; NDF, neutral detergent fiber; ADF, acid detergent fiber; L, lactic acid bacteria. P, podding stage; EF, early fruiting stage, ND, not detected.

Table 2 The chemical composition of soybean silages after 60 days of ensiling. Item

g kg-1 DM

DM

CP

EE

NDF

ADF

24.66b

12.14c

0.83c

58.84a

46.34

24.65

b

12.03

c

b

24.87

b

13.30

b

24.63

b

14.67

a

Sample 1 C L 2%M 2%ML Sample 2 C L 2%M 2%ML

1.08

bc

57.25

45.87

1.20

abc

57.53b

44.65

1.14

abc

b

57.54

46.73

26.76a

12.24c

0.95bc

58.95a

47.00

27.27

a

12.52

c

1.00

bc

a

46.77

26.65

a

13.08

b

1.33

ab

a

46.42

27.71

a

14.61

a

1.39

a

b

45.38

58.90 58.88 57.55

a-c

means in the same colomn followed by different letters differ (P<0.05).

DM, dry matter; CP, crude protein; EE, ether extract; NDF, neutral detergent fiber; ADF, acid detergent fiber.

Table 3 Significant analysis of main factors (L, M and their interaction) after 60 days of ensiling. Sample 1

Sample 2

L

M

L×M

L

M

L×M

pH

**

**

**

**

**

**

L

NS

NS

NS

NS

NS

NS

Yeast

NS

*

**

*

**

*

Coliform

**

**

**

**

**

**

Lactic acid

*

**

**

*

**

**

Acetic acid

*

*

*

*

*

*

Butyric acid

NS

**

**

NS

**

**

DM

NS

NS

NS

NS

NS

NS

CP

NS

**

**

NS

**

*

EE

*

*

**

*

*

*

NDF

*

**

*

NS

NS

*

ADF

NS

NS

NS

NS

NS

NS

NS, not significant; *, P<0.05; **, P<0.01.

Table 4 Alpha diversity of bacterial diversity at the day 0 and 60 of ensiling. Sample ID Reads OTU Chao1 Coverage Sample 1 FM 61723 177 162 0.99 C 64738 275 260 0.99 L 62837 223 206 0.99 2%M 60358 250 232 0.99 2%ML 69307 197 193 0.99 Sample 2 FM 56640 284 278 0.99 C 67345 366 344 0.99 L 51977 390 390 0.99 2%M 51569 252 233 0.99 2%ML 61522 227 226 0.99

Shannon 3.28 3.98 3.96 3.41 1.75 2.84 4.04 3.07 3.20 1.88

Fig.1. The change of pH and microbial community by plate culture during ensiling process of soybean (Sample 1).

Fig.2. The change of pH and microbial community by late culture during ensiling process of soybean (Sample 2).

Fig.3. The change of organic acid during ensiling process of soybean (Sample 1).

Fig.4. The change of organic acid during ensiling process of soybean (Sample 2).

(a)

(b) Fig.5. The PCA analysis based on pH, microbial population and organic acid in soybean silages. a, sample 1; b, sample 2.

Fig.6. The microbial community of fresh materials and silages after 60 days of ensiling revealed by NGS.

HIGHLIGHTS


Soybean ensiled with lactic acid bacteria inoculant and molasses.




All additives improved fermentation quality of soybean silage.




Combined addition of lactic acid bacteria and molasses showed best silage quality.




Molasses enriched the abundance of Lactobacillus.




Combination of lactic acid bacteria and molasses decreased Clostridia abundance.