Development and effect of a Lactobacillus plantarum inoculant on quality of maize grain silage

The main aim of these studies was the characterisation and identification of
 lactic acid (LAB) bacteria isolated from untreated silage, and the effect of
 selected bacteria (inoculant was called Silko for maize) on ensiling of
 maize high-moisture grain. Four isolates of L(L1, L2, L3 and L4) were
 characterised by the use of phenotypic assays and identified by phylogenetic
 analysis of 16S rRNA as L. plantarum. The fresh maize high-moisture grain
 was ensiled with a Silko for maize inoculant, inoculant available in the
 market (positive control) and no additive (untreated; negative control).
 After 60 days of ensiling, the results showed that the chemical composition
 and fermentation characteristics were better in treated silages with
 inoculants compared to the negative control. The contents of ash, fat and
 lactic acid (LA) were significantly higher in the silages treated with
 inoculants than in negative control. In comparison, the contents of
 cellulose, acid detergent fibre (ADF), neutral detergent fibre (NDF),
 NH3-N/total nitrogen and butyric acids (BA) were considerably lower in
 silage treated with Silko for maize compared to the positive control. The
 Silko for maize improve nutritional value and fermentation of maize grain
 silage and is a competitive product on the market.


Introduction
Preparation of quality silage is crucial for the profitability of livestock farms because it is a source of food for the periods of the year when animal nutrition is inadequate in terms of quantity and quality. Maize is a vital crop for world farmers. Maize grain is used to satisfy the energy requirement of livestock. It can be ensiled and used as an animal feed ingredient. Ensiling of moist maize grains has several advantages, such as savings on drying costs, the high nutritional value of silage, easy use. In the case of low maize grain prices on the market, the valorisation of maize is possible through animal products, and planning of the expected profit improves.
On the other hand, the preparation and storage of silage on farms is a significant problem, since it is necessary to maintain high-quality silage and achieve the maximum profitable production of milk and meat (Aragón, 2012). Silage quality directly affects feed intake and utilisation of nutrients in ruminants, as well as on milk production (Huhtanen et al., 2003). A decrease in pH prevented the loss of nutrients in silage due to higher lactic acid production (Saarisalo et al., 2007). However, when silages are exposed to the air during the opening, it leads to increase the activity of undesirable microorganisms, and that causes the decrease of dry matter and quality (Borreani et al., 2018). Also, during the ensiling process is very important that the forage mass more compacted and that the less oxygen remains in the silage. Accordingly, the inoculants are used to increase the level of lactic acid and the aerobic stability of silage over a more extended period after the opening of the silage. The use of silage inoculants (starter cultures) during silage provides a reduction in pH and the growth of undesired aerobic microorganisms (Zielińska and Fabiszevska, 2018).
The biological additives that are used for conservation of silage include saprophytic, safe bacteria within the Lactobacillus sp, which use as consortium as multiple strains (more strains inside same species), or mixed strains including different species (Jalč et al., 2009). In the world market, there are various silage bacteria inoculants for maize, among which are the frequently Lactobacillus sp. They are classified into two metabolic categories: homofermentative and heterofermentative bacteria (Contreras-Gouveia and Muck, 2006). Homofermentative bacteria produce about 90% of lactic acid and belonging to various generations of Lactobacillus including the most well-known strains of L. plantarum. These bacteria dominate inoculants products in the world. One of the reason, because it is highly competitive with epiphytic lactic acid bacteria in silage, produces large amounts of lactic acid, reduces pH and nutritional losses (Lynch et al., 2012;Đorđević et al., 2017). Also, it reduces ammonia nitrogen (Queiroz et al., 2013). In general, silage treated L. plantarum has a better fermentation quality than in untreated silage (Liu et al., 2016). However, according to Muck (2013, silage inoculants produced in cold regions may or may not be effective when used in hot regions. The purpose of this study a display of development L. plantarum inoculant (Silko for maize) and to investigate the effect on the quality of maize grain silage. The first step in the experimental work was the isolation of a large number of bacteria from different silage samples, their phenotypic and genotypic identification at the level of the species, and prepares the bacterial consortium.

Materials and Methods
Procedure for isolation bacteria. For microbiological testing, each sample was taken, about 10 g silage in 300 ml of sterile Erlenmeyer bottle of 300 ml, and 90 ml of purified water was added and incubated with stirring at 120 rpm, 30 min (Ekundayo, 2014). Subsequently, the serial decimal (from 10 times) dilution to 10 -7 was prepared in sterile phosphate buffer, pH 7.2. After that, 100 μl of each dilution was smeared on the Man, Rogosa, Sharp (MRS) agar plate (Torlak, Serbia), and were cultivated at 37°C in the anaerobic jar (BioMerieux, France), during the 72 h. The separated colonies were inoculated on MRS agar. After cultivation, it was used for phenotypic and genotypic characterization and further checked for treatment silage. All bacterial isolates were stored at 5 o C±3 o C and subculture every two weeks. For long-term storage, stock from overnight cultures was prepared and frozen in cryoprotective agents 20% glycerol and stored at -80 °C.
Preliminary phenotypic characterization. Single, clearly separated colonies were used for morphological characterisation. Initially, preliminarily tested for Gram reaction by Gram staining and catalase enzyme. The following was done sporulation, the growth temperatures range (15°C, 30°C, 37°C and 45°C); in substrates with different osmotic pressure (2%, 4% and 6.5% (w/v) NaCl), growth in aerobic and anaerobic conditions in MRS broth, during the 72h. In the next step, selected isolates were tested by standard API 50CH test, according to the manufacturer's instructions (Bio-Merieux, Montalien-Vecien, France). All experiments were done in triplicate. The isolates were further checked for inoculant on ensiling of maize high-moisture grains.
Isolation of DNA and PCR identification. Total DNA isolation from L. plantarum was done with commercial kits according to the manufacturer's protocol (BIOLINE, United Kingdom). Isolated DNA from the samples was used to identify the bacteria. Using PCR, the 16s rRNA gene was amplified using universal primers 27f (AGAGTTTGATCMTGGCTCAG) and 1492 (TACGGYTACCTTGTTACGACTT). PCR reaction was carried out in a reaction mixture of 50 μl according to the manufacturer's protocol (Thermo Fisher Scientific, USA). The PCR reactions were carried out according to the following conditions: 1. One cycle of initial denaturation 95 ºC 5min, 2. 40 cycles of denaturation 95 ºC 30s, annealing 30s 53 ºC and elongation 72 ºC 1min, 3. One cycle of final extension 72 ºC 5min.
The PCR product was checked on 2% agarose gel using electrophoresis and then purified using a commercial kit (Zymo Research, Irvine, USA) and sent to sequencing in 'MACROGEN' (Netherland) sequencing service. The sequence was bioinformatics processed using the Basic Local Alignment Search Tool (BLAST) on the NCBI.
Agar well diffusion methods. Agar-well diffusion method (AWD) was used (Harris et al., 1989) for detection of cross inhibition (antimicrobial activity) between alone strains of the genus Lactobacillus.
Preparation of silage. Maize hybrid ZP 684 (FAO 600 maturity group) was grown on a plot at the Research and Development Centre 'Agrounik' (44 ° 52 'N, 20 ° 05' E), Serbia during 2017. Preceding crop was winter wheat. Maize was sown on April 15. The sowing density was 60.000 plants ha -1 . The hybrid was harvested with a combine harvester in August when the grains had 26-32% moisture. The grains were ground in a mill to a 3-4 mm particle size. Approximately 100 kg was taken from the field and brought to the laboratory. Three treatments were used: 1. the untreated maize (negative control), 2. positive control (commercial inoculant added at 2 l t -1 of grain, contained the L. plantarum at total concentration 1 x 10 5 CFU g -1 of inoculant), 3. Silko for maize (number of colony-forming units in inoculant is 1 x 10 10 CFU ml -1 ; applied at a rate of 5 ml t -1 of grain). Maize was packed and compressed in polyethene containers volume 6l and covered with foil and a layer of sand.
Silage analysis. After 60 days of ensiling, about 450 g samples of the maize grain silage were taken from the containers for analysis. Standard methods according to AOAC (2000) were used to determine the contents of dry matter (DM), ash, crude fat (CF), crude protein (CP), acid detergent fibre (ADF) and neutral detergent fibre (NDF). Weende method was used to determine the content of cellulose, while method, according to Licitra et al. (1996) for the content of soluble nitrogen/total nitrogen. The content of NH3-N/total nitrogen was determined using a Kjeltec System 1026. A gas chromatograph (GC-2014, Shimadzu, Kyoto, Japan) was used to determine the contents of lactic-(LA), acetic-(AA) and butyric acids (BA). The pH value was measured using an electronic digital pH meter (Hanna Instruments HI 83141 pH meter).
Statistical analysis. One-Way ANOVA was used in the analysis of experimental data, using Statistical software Statistica version 10 (StatSoft, Tulsa, Oklahoma, USA). The randomized complete block analysed the trial with three replicates. Tukey's test (P ≤ 0.05) was used to compare the results.

Results
Preliminary phenotypic characterization. From 15 silage samples, 62 different colony morphologies were isolated. After the preliminary testing, 17 isolates of Lactobacillus and 2 Pediococcus were isolated. Four Lactobacillus isolates were denoted in laboratory collection bacteria as L1, L2, L3 and L4 and selected based on their high LA production. The morphological, cultural and physiological characteristics of the selected LAB performed in Table 1. According to its fermentative properties (carbohydrate substrate 49), all those which one tested Lactobacillus isolates showed the highest similarity to bacteria belonging to L. planarum / L. pentosus.

Molecular identification.
According to phenotypic and molecular characterisation (complete sequence16S rDNA isolate L1, L2, L3 and L4), the bacterial isolates were identified as L. plantarum and signed among our laboratory isolates as L. plantarum -L1, L. plantarum -L2, L. plantarum -L3 and L. plantarum -L4. Before forming a consortium of bacteria, we checked whether cross-inhibition occurs among the individual strains using the agar diffusion method. After 24 h cultivation bacteria, cross-inhibition between lactobacilli tested was not detected (Table 2).

Indicator strain
Zone of inhibition growth (mm) Test strain 0 0 0 / Values are means of triplicate determinations with standard deviations; 0: do not zone inhibition growth; / do not test Chemical composition of maize grain sample before ensiling. Chemical composition of maize grain sample before ensiling is shown in Table 3. Distinct letters in the row indicate significant differences according to Tukey's test (P ≤ 0.05); **significant at 1% level of probability, * -significant at 5% level of probability and ns -not significant.

Effect of the inoculants on maize grain silage quality.
Results showed that the ash, crude fat and crude protein were significantly higher, while ADF was significantly lower in silages treated with bacteria inoculants than in control (Table   Development and effect of a Lactobacillus …   245   4). The cellulose and NDF were significantly lower in silage treated with Silko for maize compared to positive control and negative control. The dry matter did not differ among treatments.
Fermentation characteristics of silage are influenced by inoculants ( Table  5). The values of pH, NH3-N/total nitrogen, AC and BA were lower, while LA was higher in treatments with inoculants than in negative control. The pH and AA did not differ among positive control and Silko for maize. 0.11 c 0.05 b 0.02 a 0.06 ** DMdry matter; TNtotal nitrogen; Distinct letters in the row indicate significant differences according to Tukey's test (P ≤ 0.05), **significant at 1% level of probability.

Discussion
Based on LA production, four strains of L. plantarum have been selected. Due to the absence of cross-inhibition, they are selected as a consortium of bacteria. Also, strains of Lactobacillus belong to the GRAS (General Recognized as Safety). The use of bacterial inoculants in the initial phases of fermentation in grass silage, grass-clover, alfalfa and maize aims to decrease in pH to avoid the rapid growth of harmful microorganisms and losses of dry matter and increase aerobic silage stability (Jatkauskas et al., 2013). In generally, L. plantarum inoculants used in our study improve fermentation, promoting LA production, decrease pH, NH3-N/total nitrogen, AA and BA acid contents in silage.
The higher ash and crude fat contents were recorded in silages treated with inoculants. The higher ash content is the result of the metabolism of inoculated strains of bacteria which using soluble components and thus increase the relative ash content (Đorđević et al., 2017).
Crude protein is one of the most critical animal food quality parameters so it is crucial to maintain its high level in silage. Our results showed that the silages treated with L. plantarum inoculants have significantly higher crude protein content compared to control. According to Abdul Rahman et al. (2017), addition of L. plantarum to silage increases crude protein content due to higher production of protein in the form of nitrogen content. The L. plantarum possess reductases that can reduce nitrates and nitrites to ammonia and other ammonia compounds (Rooke and Hatfield, 2003). This contributes to the increase in total protein content, although these compounds within total proteins mainly occur as non-protein nitrogen. Also, the low pH in treated silages inhibited protein degradation, as evidenced by research of Vukmirović et al. (2011).
The lowest cellulose content was recorded in maize silage treated with Silko for maize. According to Sadiya and Ibrahim (2015), the Lactobacillus sp. produces enzymes for hydrolysis of cellulose. Therefore, we can assume that the strains of L. plantarum produce these enzymes, as indicated by the research of Đorđević et al. (2017). Likewise, Koc et al. (2009) found the lowest cellulose content in sunflower silages treated with inoculant containing L. plantarum and Enterococcus faecium.
ADF and NDF were lowest in silage treated with Silko for maize. NDF did not differ between negative and positive controls. The Silko for maize increases digestibility and dry matter intake of silage and can be expected to will have a positive impact on animal performance because of the lower ADF and NDF levels in food increase animal productivity (López et al., 2018).
The pH range 3.77-4.27 indicated of well-preserved silage. The inoculants promoted the silage acidification. The low pH preserves nutrients and promotes homofermentative lactic acid bacteria in silage (Li et al., 2015). In essence, the low pH in silage reduced survival of yeasts, moulds and other undesirable silage microorganisms like Clostridia, prevent heating and spoilage silage and dry matter losses (Ren et al., 2018).
The content of NH3-N/total nitrogen was lowest in silage treated with Silko for maize. However, the NH3-N content of treated and untreated silages was <100 g/kg N which suggests successful preservation. Therefore, proteins were degraded to a greater extent in untreated silage where the pH was higher than in silage treated with Silko for maize. Thus, we can be assumed that in silage treated with Silko for maize inoculant, is the higher protein content in an intact form which animals can be utilized directly. According to Contreras-Govea et al. (2013) silage treated with L. plantarum has more true protein than untreated silage, and has positive effects on milk production.
The studied silages have the higher content of LA and lower contents of AA and BA than reference values for high-quality silage (> 6.5%, < 3-4% and < 0.5%, respectively), indicating proper lactic acid fermentation. According to Shaver (2003), good silage contains from 65 to 75% LA. In our research, all silages have satisfactory LA content. The highest LA and lower BA levels were recorded in silage treated with Silko for maize. The highest AA level was recorded in untreated silage. Generally, the increase in LA content in silage decreased pH and prevents secondary fermentation. The low pH inhibits clostridia growth, resulting in lower BA content in investigated silages. Consistent with our findings, Muck (2013) concluded that the homolactic bacteria have greater efficiency of glucose utilisation, produced higher LA in silage, reduced pH, and thus prevent undesirable microbes. According to Kung and Shaver (2001), the ratio of LA to AA in the silage should be of more than 3:1, indicating the excellent fermentation. In our case, the addition of inoculants increased this ratio compared to control silage and indicated very good fermentation.