Ensiling is a preservation method for moist forage crops. It is based on solid-state lactic acid fermentation under anaerobic conditions whereby lactic acid bacteria (LAB) convert water-soluble carbohydra-tes (WSC) into organic acids, mainly to lactic acid. As a result the pH decreases, and the moist crop is preserved (Weinberg and Ashbell, 2003). A moist crop can support the growth of a wide range of mi-croorganisms, most of which will degrade the crop’s nutrient value to livestock. Ho-wever, ensiling generally controls micro-bial activity by a combination of an ana-erobic environment and a natural fermen-tation of sugars by lactic acid bacteria (LAB) on the crop (Muck, 2010).
Legumes are difficult to ensile successfully without an additive (Carpintero et al., 1969; Ohshima et al., 1979). This is espe-cially true for alfalfa (Medicago sativa L.). McDonald et al. (1991) considered that the difficulties in ensiling legumes were attri-butable to three factors. Firstly, they are highly buffered; secondly, they tend to have low water-soluble carbohydrate (WSC) contents; and thirdly, they often have low dry-matter (DM) content.
The number of LAB present on alfalfa plants at harvest may be too low to ensure rapid and efficient preservation, and the-refore silage inoculants have been deve-loped to improve the nutritive value of si-lages and to reduce risks during ensiling (Čabarkapa, 2010a).
The possible beneficial effect of LAB ino-culants depends on their composition, concentration and the properties of the crop being ensiled (Tengerdy et al., 1991).
Regarding the effect of bacterial inoculant, some authors (Weinberg et al., 1988; Ten-gerdy et al., 1991; Masuko et al., 2002) indicated that inoculation of the alfalfa si-lages with LAB enhanced the silage qua-lity, while others reported no effect (Lind-gren et al., 1983; Ohyama et al., 1973). The aim of this study was to determine effects of adding a silage inoculant with a new combination of LAB strains to alfalfa, grown under south-eastern European conditions, on fermentation dynamics and aerobic stability of silage.

Fifth-cut alfalfa (cultivar Banat) was har-vest in 2010 in province of Vojvodina, Serbia, using a precision silage chopper to obtain a theoretical 20 mm chop length. Investigated inoculant Bonsilage Forte (Schaumann, Agri Austria GmbH) contains combination of homo-fermentative strains of Lactobacillus paracasei (DSM 16245), Lactobacilus lactis (NCIMB 30160) Pedio-coccus acidilactici(DSM 16243). It was used as an inoculant by mixing 0.30 g with 0.1 dm3 of water (to provide2.5×105 co-lony forming units of LAB per gram of fresh maize), and spraying over 30 kg of alfalfa. The application rate was in accor-dance with the level of LAB in the ino-culant as determined bythe manufac-turer. In order to add the same amount of moisture as in the treated alfalfa, the control wastreated by spraying 0.1 dm3 of water on 30 kg alfalfa.
The alfalfa (280.6 g/kgDM) was compact-ted into 1.8 dm3 transparent containers-with a special water valve to enable gas release, as proposed by Palić etal. (2011). Each container was filled with approximately 0.65 kg (wet mass) alfalfa without a headspace, and a packing density of approximately 360 kg/m3 was obtained.
A total of 12 containers were divided into two groups. Each group consisted of three control containers, without inoculant, and three inoculated samples, with Bonsilage Forte added. Containers were stored in dark roomat a temperature of 21-25 0C.
On day 0, fresh alfalfa was collected for subsequent chemical analysis. First group was opened on day 14 and second on day 49 day for determination of pH, dry matter (DM), crude protein (CP), ammonia nitro-gen (NH3-N), lactic acid (LA) and volatile fatty acids (VFA). The pH, DM and CP were determined following the procedures of AOAC (2005). The NH3-N, LA acid and VFA were determined according to stan-dard methods (Die Chemische Untersu-chung von Futtermitteln, 1997). On day 49 silages were also subjected to an aerobic stability test conducted according to pro-cedure of Ashbell et al. (1991).
Statistical Analysis System STATISTICA (2011) was used for analyzing variations (ANOVA) and least significant differences. The level of significance was set at p<0.05.

Dry matter and crude protein content of the fresh alfalfa and silage samples is pre-sented in the table 1. It can be seen that there was no significant difference bet-ween control (C) and inoculated samples (BF).
Initial pH value of alfalfa (6.43) decreased to a greater extent in inoculated samples and was about 5.0 on day 49, compared to control samples where it was slightly be-low 6.0 (Figure 1). Also, pH drop of inocu-lated silage in the first days of ensiling was faster, which is important for conserving of nutrients of the silage by inhibiting spoil-age microorganisms (present in the natu-ral plant microflora) and hetrofermentative bacteria (enterobacteria) and promoting homofermentative lactic-cid bacteria. Va-lue of pH is a key criterion to evaluate silage fermentation. Generally, the lower the pH, the better preserved and more stabile is the silage (Seglar, 2003).
Fermentation parameters of experimental silages (lactic acid, acetic acid, butyric acid and ammonia nitrogen concentration) are shown in the table 2. Concentration of lactic acid was significantly higher in ino-culated samples and its concentration was doubled on day 49 compared to day 14. Generally, the presence of high lactic acid levels indicates efficient fermentation and minimal dry matter loss (Seglar, 2003).

C - without inoculant
BF - with inoculant
Results are given as mean ± standard deviation
abMeans with different superscripts in the same column are significantly different (P<0.05)

C - without inoculant
BF - with inoculants
Results are given as mean ± standard deviation
abMeans with different superscripts in the same column are significantly different (P<0.05)

Acetic acid concentration was significantly higher in inoculated samples which have positive effect on silage quality since this acid is important for inhibition of growth of moulds and yeasts (McDonald et al., 1991). The effect of acetic acid on fungal growth is related to the undissociated con-centration in solution. Thus a given con-centration of acetic acid becomes more inhibitory to yeasts and molds as silage pH decreases (Čabarkapa et al. 2010b).
Presence of butyric acid is the result of Clostridial activity (Seglar, 2003). Butyric acid was detected only in control samples. Clostridia can grow at lower pH values than enterobacteria making them more di-fficult to control (Čabarkapa et al. 2010b). If pH value cannot be decreased deep enough the Clostridia spores germinate and degrade lactic acid to butyric acid.
Concentration of ammonia nitrogen (NH3-N) increased through the expe-riment, but it was significantly lower in ino-culated samples. Higher ammonia indic-ates protein brake down from proteolytic enzymatic activity (Seglar, 2003).
Results of aerobic stability test on day 49 (Figure 2), expressed by released CO2 upon exposure to air, showed positive effect of inoculants on aerobic stability of silage. Amount of released CO2 was three times lower in inoculated samples com-pared to control.

In this investigation, rapid drop of pH occurred in the first days of ensiling and it was much stronger in inoculated samples. Value of pH continued to decrease through whole period of ensiling for both control and inoculated samples, and rea-ched much lower value in inoculated sam-ples.
Lower pH in inoculated samples probably inhibited protein degradation and therefore concentrations of ammonia-nitrogen were lower in those samples, which demonstra-tes positive effect of Bonsilage Forte on nutritive value of silage.
Addition of inoculants resulted in a signify-cantly higher lactic concentration than in the control samples. Lactic acid is the strongest of all silage acids and its pre-sence drops pH more effectively than other volatile fatty acids.
Absence of butyric acid in inoculated samples confirms that Clostridial second-dary fermentation did not take place and that Bonsilage Forte acted efficiently against Clostridia.
The aerobic stability of silage, as measu-red by the release of CO2 after exposure to air, was significantly improved by addi-tion of BF.

This paper is a result of the research within the project III046012 “Study of mo-dern biotechnological methods in the pro-duction of animal feed in order to increase competitiveness, quality and safety of the feed”, financed by the Ministry of Educa-tion and Science, Republic of Serbia.