Skip to main content
  • Original Article
  • Open access
  • Published:

Lactic acid bacteria diversity in corn silage produced in Minas Gerais (Brazil)



The diversity of lactic acid bacteria (LAB) in silages produced in warm climate countries is not well known. This study aimed to identify and characterise the metabolic and genotypic aspects of autochthonous LAB isolated from corn silage produced in the state of Minas Gerais, Brazil.


Eighty-eight LAB were isolated. To evaluate their performance at the strain level, all isolates were distinguished among strains using random amplified polymorphic DNA polymerase chain reaction (RAPD-PCR) and repetitive extragenic palindromic PCR (REP-PCR) techniques. The organic acid and ethanol production were determined by high-performance liquid chromatography (HPLC).


The fingerprints obtained by RAPD-PCR with a M13 primer were more discriminatory than those obtained with the REP-PCR technique using a (GACA)4 primer. Moreover, 28 representative isolates were identified as Lactobacillus acidophilus, L. buchneri, L. casei, L. diolivorans, L. hilgardii, L. paracasei, L. parafarraginis, L. plantarum, L. rhamnosus, L. zeae and Pediococcus acidilactici. Different fingerprinting profiles between isolates within the same species were observed. However, some strains isolated from different silages showed the same band profile, thus suggesting the presence of clusters with high similar fingerprints in silages from various regions.


A variation in LAB diversity was observed in the silages of the evaluated regions, with L. rhamnosus and L. buchneri showing the highest distribution. Differences in organic acid production were observed among the strains belonging to the same species. This research contributes to a better understanding of the LAB community present in corn silage produced in warm climates. These strains will be studied as potential silage starters.


Corn (Zea mays L.) has adequate characteristics for silage fermentation and is the forage often used in many parts of the world (Bernardes and do Rêgo 2014; Khan et al. 2015). Due to the nutritional quality, this forage has been extensively used to optimise the yield of milk production in order to meet the increasing demand for dairy products (Yang et al. 2019). The use of corn silages has increased in several countries such as Brazil, Mexico (South and Central America) (Bernardes and do Rêgo 2014; Prospero-Bernal et al. 2017), Thailand, Malaysia (Asia) (Ohmomo et al. 2002; Khaing et al. 2015) and Kenya and Ethiopia (Africa) (Tamir et al. 2012; Makau et al. 2018).

Typically, many aerobic and anaerobic microorganisms are found in silage and affect its quality. As the anaerobic atmosphere inside the silo increases, the population of various facultative anaerobic microorganisms also rises, primarily fermenting sugars into organic acids (lactic, acetic, propionic, butyric, formic, caproic and valeric acids), which in turn exhibit antimicrobial activity preventing the spoilage by undesirable microorganisms such as yeasts and filamentous fungi (Borreani et al. 2018; O’Brian et al. 2007). Moreover, the impact of the presence of filamentous fungi can increase if toxic secondary metabolites (mycotoxins) are produced. At the time of ensiling, plant material encloses a large range of microorganisms (Fabiszewska et al. 2019), but the fermentation process favours the multiplication of lactic acid bacteria (LAB). LAB are part of the main group of microorganisms that act in the silage fermentation process. LAB include bacteria from a number of genera, such as Enterococcus, Lactobacillus, Lactococcus, Leuconostoc, Pediococcus, Weissela and Streptococcus (Vandamme et al. 2014). Of these, Pediococcus, Lactococcus, Enterococcus and Weissela are important in the initial stage of fermentation because they ensure the environment remains acidic, which becomes predominantly colonised by the Lactobacillus genus (Lin et al. 1992).

Environmental factors such as temperature, humidity and precipitation seem to have influence in the microbial population in silages, especially in warm and humid areas (Muck, 2013). During the ensiling process, rainfall and humidity have impact in the epiphytic bacterial community, and temperature affects the richness of bacterial species and subsequent silage fermentation. Fortunately, these factors seem to have negative impact in some undesirable microorganisms and do not affect the Lactobacillus and Leuconostoc genus (Guan et al. 2018).

It is generally recognised that changes in the LAB population diversity occurred during the ensiling process in which an adequate ensilage usually starts with homofermentative LAB species quickly growing after sealing, later being replaced by heterofermentative LAB species when substrate availability becomes limited (McDonald et al. 1966; McDonald et al. 1991; Yang et al. 2006). Furthermore, higher ensiling temperatures also contribute to accelerate the shift from a homolactic to a heterolactic microbial population (Bernardes et al. 2018).

The identification of autochthonous strains and the study of LAB in silage produced under different conditions are necessary to improve silage quality. LAB identification is an important step, but its fermentation performance is always strain dependent. Therefore, it is of primary importance to use techniques capable of discriminating among strains. Typing methods might be used to trace individual strains (Abdollahniya et al. 2018). Several studies have reported success using random amplified polymorphic DNA polymerase chain reaction (RAPD-PCR) for the differentiation of LAB isolates (Rossetti and Giraffa 2005; Parente et al. 2016; Abdollahniya et al. 2018). Repetitive extragenic palindromic PCR (REP-PCR) is another powerful tool used to identify several types of LAB, including Lactobacilli (Gevers et al. 2001; Dolci and Cocolin 2017). These techniques can be used to rapidly compare a large number of isolates and reveal relationships at the species and subspecies levels. Despite the importance of culture-independent techniques, classic studies that use the recovery of microorganisms in culture media remain necessary.

To the best of our knowledge, there is no report in the literature showing the isolation and identification of LAB present in farm scale corn silage produced in warm climates. Thus, the present study aimed to identify the LAB isolated from corn silage produced in farms from different regions of the state of Minas Gerais, Brazil, using a genotypic and phenotypic (organic acid and ethanol production) approach.

Materials and methods

Sampling and bacterial isolation

LAB strains used in this study were isolated from samples of corn silages produced in bunker silos in 9 Brazilian micro-regions of Minas Gerais State: Lavras (21° 14′ S, 44° 59′ W, 919 m), Elói Mendes (21° 36′ S, 45° 33′ W, 907 m), Silvianópolis (22° 01′ S, 45° 50′ W, 897 m), São Sebastião do Paraíso (20° 55′ S, 46° 59′ W, 991 m), Passos (20° 43′ S, 46° 36′ W, 745 m), Muzambinho (21° 22′ S, 46° 31′ W, 1048 m), Piranguinho (22° 24′ S, 45° 31′ W, 837 m), Itumirim (21° 19′ S, 44° 52′ W, 871 m) and Nepomuceno (21° 14′ S, 45° 14′ W, 840 m) (Santos 2016). The climate in these regions is classified as Cwa (warm temperate climate with dry winter), the precipitation occurs mainly from October to March and the average temperatures in the winter and summer are 19.7 °C and 22.7 °C, respectively (Sá Júnior et al. 2012). Corn silage samples were collected from 54 dairy farms, six from each region. For the bacterial isolation experiment, silage samples were selected according to fermentative characteristics of pH and temperature. The total count of yeast, filamentous fungi and bacteria (Enterobacteriaceae and aerobic spore-forming bacteria) were also taken into consideration (Santos 2016; Santos et al. 2020). In this sense, two samples from each region, one of the farms that showed overall the best (the lowest pH, silage temperature, Enterobacteriaceae, aerobic spore-forming bacteria and yeast and filamentous fungi population) and one of the farm that showed the worst characteristics (opposite to the previously mentioned) were selected, making a total of 18 samples.

A total of 122 bacterial strains were isolated and cultivated at 37 °C for 48 h in MRS agar medium (CM0361, Oxoid; Hampshire, England). The isolates were preserved and maintained in frozen stocks at − 80 °C in MRS broth medium (CM0359, Oxoid) with glycerol 20% (v/v) as cryoprotective agent.

Genomic DNA extraction

Isolates were reactivated and 10 μL of active broth culture was re-inoculated into 2 mL MRS broth plus cysteine (0.5 g/L) for 24 h at 37 °C. Post incubation, the bacteria cells were harvested by centrifugation at 10,000 rpm for 10 min at 4 °C. The pellet was used for genomic DNA extraction that was performed according with a protocol previously described by Sachinandan et al. (2010) using phenol-chloroform.

PCR fingerprinting amplification

Genomic DNA obtained from 105 different strains was used as a template for PCR fingerprinting using the M13 primer with sequence 5′-GAGGGTGGCGGTTCT-3′ and (GACA)4 primer with sequence 5′-GACAGACAGACAGACA-3′. PCR reactions were performed in 25 μL reaction mixtures containing NZYTaq 2x Green master Mix (Nzytech; Lisbon, Portugal), genomic DNA (50 ng) and 0.4 mM of primer. Amplification conditions for primer M13 was according Torriani et al. (1999). For primer (GACA)4, amplification reactions were performed according to the optimised protocol previously described by Shehata et al. (2008).

Amplified products were resolved by electrophoresis (50 V for 1.5 h) on 0.8% (w/vol) agarose in 0.5X TAE buffer gels, stained with GreenSafe Premium (MB13201, Nzytech; Lisbon, Portugal) and photographed with a Molecular Imager® ChemiDoc™ XRS+ Imaging System (170-8070; BioRad Laboratories, CA, USA). One kbp NZYDNA ladder III (MB04402, Nzytech; Lisbon, Portugal) was used as a DNA molecular weight marker and as a normalisation reference.

Identification to species level by 16S rRNA gene sequencing

A total of 28 representatives isolates were selected from the RAPD-PCR clustering and analysed for 16S rRNA gene using the primers 27F (5′-AGAGTTTGATCCTGG CTCAG-3′) and 1492R (5′-GGTTACCTTGTTACGACTT-3′) (Lane, 1991). Amplification reactions were performed in a total volume of 50 μL.

Amplified PCR products were sent to sequencing at STABVIDA (Caparica, Portugal). The sequences were then compared with the Gen-Bank database using the BLAST algorithm (National Center for Biotechnology Information, Maryland, USA).

Characterisation of organic acid and ethanol production

To determine the organic acid and ethanol produced in the fermentation, the 88 strains were cultivated in MRS broth for 48 h at 37 °C. After this period, the inoculum was standardised using a spectrophotometer at 620 nm wavelength at an optical density of 1.0. Subsequently, 200 μL of each strain was inoculated into 1.8 mL of MRS broth, which was incubated at 37 °C. After 48-h fermentation, samples of the cultures were analysed for lactic acid, acetic acid, propionic acid and ethanol, using high-performance liquid chromatography (HPLC) (JASCO model LC 4000; (JASCO International Co., Tokyo, Japan). The acids and ethanol were identified by comparing their retention times with those of known standards. The HPLC apparatus was equipped with a dual-detection system consisting of a UV detector (JASCO UV-2075Plus) and a refractive index detector (JASCO RI-2031). An ion exclusion column from BioRad (Aminex HPX-87H; 7.8 mm i.d., 30 cm long) operated at 60 °C was used for the chromatography separation. The mobile phase consisted of 0.005 M sulphuric acid solution with a flow rate of 0.7 mL/min. The acids were detected by UV absorbance. Ethanol was identified using the refractive index detector.

Data analysis

Pictures of the obtained gels were analysed using BioNumerics software (version 6.6; Applied Maths BVBA, Sint-Martens-Latem, Belgium). Calculation of similarity of the PCR fingerprinting profiles were based on the Pearson correlation coefficient. A dendrogram was deduced from the matrix of similarities by the unweighted pair group method using arithmetic average (UPGMA) clustering algorithm. Data regarding the production of metabolites by strains were analysed using principal component analysis (PCA) using the XL Stat software, version 7.5. The CORR procedure of SAS 9.3 (SAS Institute 2012) was used to analyses the correlation between the data of silage characteristics as dry matter (DM), pH, lactic, acetic, propionic and butyric acids, ethanol, 1,2-propanediol and storage time (Santos et al. 2020) and LAB species identified in the silages.


Genotypic characterisation

From the total of isolates, 88 were identified as LAB and 17 isolates as non-LAB. Eleven different species of LAB were identified (Table 1). These species belong to four separate phylogenetic groups of the genus Lactobacillus (L. delbrueckii, L. buchneri, L. casei and L. plantarum) and the genus Pediococcus. Lactobacillus buchneri, and L. rhamnosus were found to have a broad distribution among the silage samples; they were isolated in four different regions. In contrast, L. parafarraginis, L. zeae, L. acidophilus and L. casei were isolated in only one region. Samples of corn silage collected in the Itumirim and Nepomuceno regions showed the highest diversity of LAB, and the largest number of identified LAB (23 strains) were isolated from samples collected in the first region; P. acidilactici, L. paracasei, L. rhamnosus and L. plantarum were isolated in both regions. Samples collected in the Silvianópolis, São Sebastião do Paraíso and Muzambinho regions showed the least diversity of LAB; only L. buchneri, L. rhamnosus and L. diolivorans, respectively, were isolated.

Table 1 Distribution of LAB isolates and corn silage characteristics according to the collection regions of samples

The 88 isolates subjected to DNA fingerprinting generated bands ranging in size from 200 to 3000 bp. The fingerprints obtained by RAPD-PCR with a M13 primer (Fig. 1) were more complex and discriminatory than those obtained with the REP-PCR technique using a (GACA)4 primer (Fig. 2), being possible to group the fingerprint profiles into well-defined clusters using the first approach, while with REP-PCR technique, it was not possible to obtain a quality grouping. The average reproducibility of the M13 primer technique was 96%. This value was used to estimate the differentiation among the isolates (Fig. 1). All the isolates were clearly grouped into separate clusters according to their respective taxonomic designations, with eleven clusters detected (Fig. 1). Based on the grouping by M13, 28 representative isolates were selected and sequenced. The isolates of clusters 1, 3, 4, 5, 6, 7, 8, 10 and 11 were identified as Pediococcus acidilactici (7 isolates), Lactobacillus paracasei (11 isolates), L. rhamnosus (10 isolates), L. diolivorans (10 isolates), L. parafarraginis (2 isolates), L. hilgardii (17 isolates), L. plantarum (6 isolates), L. casei (5 isolates) and L. zeae (4 isolates), respectively. In the cluster 2, L. paracasei (2 isolates) and L. acidophilus (1 isolate) were grouped with 95.4% similarity. Cluster 9 contained 12 isolates of L. buchneri and 1 isolate of L. casei, was grouped with 94.6% similarity.

Fig. 1
figure 1

Dendrogram generated after cluster analysis of fingerprints using M13 primer of the UFLA SLM and CCMA LAB isolates. Stars indicate isolates selected to sequence analysis. Collection regions of corn silage samples: Lavras (1), Elói Mendes (2), Silvianópolis, (3), São Sebastião do Paraíso (4), Passos (5), Muzambinho (6), Piranguinho (7), Itumirim (8) and Nepomuceno (9). Superscripts codes after species names correspond to the phylogenetic groups: 1Pediococcus group; 2Lactobacillus casei group; 3L. delbruecki group; 4L. buchneri group; 5L. plantarum group

Fig. 2
figure 2

Dendrogram generated after cluster analysis of REP-PCR fingerprints using (GACA)4 primer of the UFLA SLM and CCMA LAB isolates. Stars indicate isolates selected to sequence analysis. Collection regions of maize silage samples: Lavras (1), Elói Mendes (2), Silvianópolis, (3), São Sebastião do Paraíso (4), Passos (5), Muzambinho (6), Piranguinho (7), Itumirim (8) and Nepomuceno (9). Superscripts codes after species names correspond to the phylogenetic groups: 1Pediococcus group; 2Lactobacillus casei group; 3L. delbruecki group; 4L. buchneri group; 5L. plantarum group

Isolates of P. acidilactici and L. rhamnosus were grouped with 94.2% and 96.1% similarity, respectively. The L. diolivorans, L. hilgardii and L. parafarraginis isolates were grouped with 95.7% similarity. The isolates of L. plantarum and L. zeae were grouped with 96.8% and 98.4% similarity, respectively.

The presence of L. diolivorans was associated with silages with high concentrations of DM (r = 0.562), acetic (r = 0.487) and propionic (r = 0.526) acid and long storage time (r = 0.469) (Table 2). Positive correlation was observed between the presence of L. buchneri and L. hilgardii (r = 0.475) in silages. Lactobacillus zeae was more frequently isolated in silages with high pH (r = 0.891) and 1,2-propanediol (r = 0.617) concentration and in silages with low acetic (r = − 0.538) and propionic (r = − 0.474) acid concentrations. The presence of L. casei was strongly correlated with L. plantarum (r = 0.866) and L. acidophilus (r = 0.889) species.

Table 2 Pearson correlation coefficients between corn silage characteristics and lactic acid bacteria isolated from these silages

Organic acid and ethanol production

Differences in the production of acids and ethanol were observed between the analysed isolates (Table 3). Isolates classified within the same carbohydrate fermentation patterns and of the same species showed different fermentative characteristics. The CCMA 0770 (L. hilgardii) strain showed the highest production of lactic acid (25.33 g/L) and produced 2.55 g/L of acetic acid. The strain UFLA SLM 075 (L. diolivorans) showed the highest production of acetic acid (4.11 g/L), produced 16.8 g/L lactic acid and showed the lowest lactic acid/acetic acid ratio. The highest ratio of lactic acid/acetic acid was 21.8 (L. paracasei). Fifty strains produced less than 0.6 g/L ethanol and the other 38 strains produced between 1.17 and 4.64 g/L, these being predominantly obligately heterofermentatives. The UFLA SLM 223 (P. acidilactici) did not produce any detectable amounts of ethanol or acetic acid, while the UFLA SLM 17 (L. hilgardii) and CCMA 0785 (L. casei) isolates did not produce detectable ethanol. None of the analysed isolates produced detectable amounts of propionic acid.

Table 3 Contents of lactic, acetic, and ethanol by CCMA and UFLA SLM LAB isolated from corn silage

In the principal component analysis (PCA), the first two components (PC1 and PC2) explain 93.27% of the total variance (Fig. 3). Some of the isolates only correlated with lactic acid (left upper quadrant). The isolates that correlated with acetic acid (right upper quadrant) also correlated with ethanol. The isolates shown in the lower quadrants produced smaller amounts of both acids and did not correlate with any of the metabolites.

Fig. 3
figure 3

Principal component analysis of the lactic acid, acetic acid and ethanol of UFLA SLM and CCMA isolates evaluated

The differences in organic acid production among the 88 isolates were due to the metabolism of the bacteria. In the isolates with a homofermentative metabolism, the average lactic acid content was 18.89 g/L, which is higher than the average content observed in isolates with an obligately heterofermentative metabolism (14.96 g/L) (Table 3). The average acetic acid content (2.41 g/L) and ethanol content (2.98 g/L) was relatively higher in the isolates that have an obligately heterofermentative metabolism in comparison with those with a facultatively heterofermentative metabolism (1.44 g/L and 2.41 g/L) and a homofermentative metabolism (1.21 g/L and 0.27 g/L), respectively.


All the LAB species found in the evaluated silages were previously associated with silages from corn or other plants produced at different locations, including warm and tropical regions (Parvin and Nishino 2009; Doi et al. 2013; Zhou et al. 2016). Of the 11 isolated species two presented homofermentative metabolism, three obligately heterofermentative metabolism and six species with facultatively heterofermentative metabolism. Heterofermentative bacteria generally resist longer during silage and dominate fermentation over longer storage periods (Carvalho et al. 2016; Blajman et al. 2018). In this study only Lactobacillus diolivorans (obligately heterofermentative) was correlated with long storage time, all silages presented average fermentation times over 65 days. Zhou et al. (2016) examined the influence of ensiling temperature (5 to 25 °C) on fermentation characteristics and microbial counts of corn silage and the results clearly demonstrated that changes in the LAB population diversity occurred during the ensiling process. An expressive increase in LAB counts was detected after 28 days of ensiling at 25 °C. Furthermore, in these silages, the growth of L. plantarum and P. pentosaceus was gradually replaced by L. buchneri, whereas the homolactic bacterium L. sakei and L. curvatus dominated the fermentation at 5 and 10 °C, supporting the broad consensus concerning the shift from homo- to heterofermentative LAB species during ensiling. The correlation between L. buchneri and L. hilgardii (both obligately heterofermentative) presence in silages was observed. However, there was no correlation of the presence of these species with long silage storage time.

The isolation of L. acidophilus in silages is not common and was isolated only in the Itumirin region. In this region were observed great LAB species diversity, and great population of filamentous fungi and yeasts. Lactobacillus acidophilus with feruloyl esterase activity (FAE) was isolated from corn silage produced in laboratory silos in Japan. LAB that show activity with this enzyme can improve silage quality and digestibility (Xu et al., 2017a). Strains of this species may have probiotic characteristics (Bomba et al., 2012). In review, Queiroz et al. (2018) report L. acidophilus reduced the pathogenic Escherichia coli presence in the intestine when orally administered to ruminants. The presence of L. acidophilus was positively correlated with L. plantarum and L. casei in silages, although these microorganisms has been used as starter cultures in several fermentative processes, we have not found any reports in the literature about the possible explanation for this association.

Lactobacillus buchneri is one of the obligately heterofermentative LAB often studied as a silage inoculant in recent years (Blajman et al., 2018). This species was frequently isolated in corn silages produced in southern Minas Gerais, Brazil. Rossi and Dellaglio (2007) and Ávila et al. (2009) also found a high L. buchneri population in alfalfa, corn (Italy) and sugarcane (Brazil) silages. Because this species is obligately heterofermentative and some strains present a high production of acetic acid, the main result of its inoculation in silages is the reduction of the yeast and mould population and improvement in aerobic stability (Blajman et al. 2018). Lactobacillus buchneri can produce acetic acid by hexoses fermentation or by the anaerobic degradation of lactic acid. During the anaerobic degradation of lactic acid, 1,2-propanediol is also produced. Lactobacillus diolivorans, a LAB isolated from corn silage (Krooneman et al., 2002), is associated with the degradation of 1,2-propanediol to propionic acid and 1-propanol. The correlation of the L. diolivorans presence with silages with high propionic acid concentration may be justified by this metabolism. The products of L. diolivorans metabolism are important for improving silage aerobic stability. There are few studies that describe the presence of this species in silages and few studies on the inoculation of L. diolivorans in silages; therefore, how this bacterium acts during the silage fermentation is still unknown.

The species L. hilgardii was isolated in the silages sampled in the regions of Elói Mendes and Piranguinho, and in this last region only obligately heterofermentative species were identified. Lactobacillus hilgardii was isolated during the fermentation of sugarcane silage in experimental silos (Ávila et al. 2014). The inoculation of the CCMA 0170 (L. hilgardii) strain in sugarcane and corn silages resulted in lower dry matter loss, lower yeast and filamentous fungi population and higher acetic acid concentration and aerobic stability (Assis et al. 2014; Carvalho et al. 2015; Reis et al. 2017). The inoculation of the CNCM I-4705 (L. hilgardii) strain on corn silage produced in Italy obtained similar results (Ferrero et al. 2019). The L. parafarraginis species, also belonging to the L. buchneri group, was isolated only in the Lavras region. The silage characteristics (DM, pH, acid concentrations, microorganism’s population) of this region can be classified as of good quality and in these silages was not detected filamentous fungi growth. This species has recently been studied as a silage inoculant as reviewed by Muck et al. (2018). Inoculation with L. parafarraginis has resulted in silages with higher aerobic stability when compared with control silage or silage inoculated with L. buchneri when fermentation occurred at low temperatures (Liu et al. 2014). The diversity of LAB inoculated with L. brevis (SDMCC050297) and L. parafarraginis (SDMCC050300) in corn stover silages in China was lower when compared with the non-inoculated silages. Lactobacillus plantarum and Lactococcus lactis were dominant in control silages, and L. brevis and L. parafarraginis were dominant in the silages inoculated with these species after 10 days of storage (Xu et al. 2017b). These authors observed the amount of LAB and the abundance of Lactobacillales in inoculated silages were higher than in the control.

Lactobacillus plantarum was isolated in only two regions, although the presence of this species in silages is common everywhere (Parvin and Nishino 2009; Ávila et al. 2014; Carvalho et al. 2016; Hu et al. 2018). The use of L. plantarum as an inoculant is well known and new strains are constantly evaluated, making this species one of the most studied as an inoculant for silage (Blajman et al. 2018; Hu et al. 2018). The primary results of L. plantarum inoculation in silages are a fast drop in pH and lower DM losses. However, inoculation results in decreased aerobic stability, and therefore. this species is widely used in combination with other LAB species, mainly obligately heterofermentative species.

The facultative heterofermentative L. casei species has been studied and used as a silage inoculant (Muck et al. 2018). The occurrence of this species in the silages evaluated was low (found in only one region). The silages of this region had on average the highest filamentous fungi and yeast counts and LAB diversity. In these silages were also observed lowest average of propionic acid content (6.6 g/kg DM) that has antifungal effect. The safety and efficacy of the L. casei (DSM 28872) strain used as an additive intended to improve ensiling was analysed, and the results showed this strain has the potential to improve silage production by reducing dry matter loss and enhancing protein preservation (Rychen et al. 2017). Other species belonging to the L. casei group that were isolated are L. paracasei, L. rhamnosus, and L. zeae. The use of L. paracasei (Avila et al. 2010) and L. rhamnosus (Li and Nishino 2011) as additives for silage was reported; however, when this species was compared with other LAB, the results of the inoculation were not satisfactory. Lactobacillus zeae was found only in the Passos region, the silages from this region presented high pH values and low lactic and acetic acid concentrations. This species was also isolated from the alfalfa silage (which also have high pH values) produced in farm scale silos in Italy (Rossi and Dellaglio 2007). Little is known about the performance of this species during silage fermentation, and no studies evaluating L. zeae as an inoculant were found.

Pediococcus acidilactici was identified in three evaluated regions. This species is among the most common homofermentative LAB used as silage inoculants, although there are few studies evaluating its performance and it is frequently inoculated together with other LAB (Blajman et al. 2018). Pediococcus acidilactici dominate the early stages of ensilage and rapidly decrease pH (Fitzsimons et al., 1992). Although no significant correlation was observed with storage time, this species was isolated from silages with an average storage time of 380 days.

The ability to clearly differentiate between the LAB isolates was low for the isolates identified using the (GACA)4 primer. A reduced number of bands were observed, making it difficult to distinguish single groups according to species. In contrast, the M13 fingerprint profiles were relatively heterogeneous. This variability can be useful for intraspecific differentiation of isolates within the same species (Turková et al. 2012). The M13 primer has already been used for Lactobacilli typing and has proven useful for the differentiation of a wide range of Lactobacilli at the species and subspecies levels, and potentially, up to the strain level (Rossetti and Giraffa, 2005; Švec et al. 2010). In evaluating the genotypic profile obtained by M13, small differences in the band profile between isolates within the same species were observed. However, some strains isolated from different silages showed the same band profile, thus suggesting the presence of clusters with high similar fingerprints in silages from various regions. The identification at a species level substantiates suitability of bacterial strains for industrial application. For example, their performance as starter cultures, which is strain dependent. The results demonstrate the usefulness of M13 fingerprinting as a rapid method and alternative strategy for subspecies identification; this technique enabled the precise grouping of isolated LAB isolates.

Knowledge about LAB metabolism is important to predict its effect as a starter on silage. Lactic acid plays an important role in the rapid reduction of pH, causing the inhibition of undesirable microorganisms that grow in the anaerobic fermentation stage, but it does not act by inhibiting the microorganisms’ aerobic deterioration. In addition to contributing to pH reduction, acetic acid and propionic acid have an antifungal effect and inhibit spoilage during the aerobic stage of silage fermentation.

The results of this study show that, in general, bacterial species with a heterofermentative metabolism produce higher concentrations of acetic acid, which is important for inhibiting aerobic spoilage. This property is desirable in silage with a spoilage problem, such as a corn silage. However, it is important to note that there may be differences between strains of the same species. Thus, the selection of strains for inoculants should be based at the strain level and not only the species.


This research has contributed to our understanding of the bacterial community structure of LAB present in farm scale corn silage produced in a warm climate, such as the Brazilian state of Minas Gerais. Eleven different LAB species were identified and characterised. All the LAB species found were previously associated with silages from corn or other plants produced at different locations. The species L. buchneri, L. diolivorans, L. hilgardii, L. paracasei and L. rhamnosus were found in greater proportion in the silages evaluated. Differences in organic acid production were observed among the strains belonging to the same species. Further studies can assess the potential of these isolates as starter cultures to produce corn silage and other forage crops.


Download references


The study was financially supported by the Brazilian agencies CNPq (Conselho Nacional de Desenvolvimento Científico e Tecnológico), FAPEMIG (Fundação de Amparo a Pesquisa do Estado de Minas Gerais) and CAPES (Coordenação de Aperfeiçoamento de Pessoal de Nível Superior). At the University of Minho the study was supported by the Portuguese Foundation for Science and Technology (FCT) under the scope of the strategic funding of the UID/BIO/04469/2013 unit and COMPETE 2020 (POCI-01-0145-FEDER-006684), and the BioTecNorte operation (NORTE-01-0145-FEDER-000004) funded by the European Regional Development Fund under the scope of Norte 2020—Programa Operacional Regional do Norte.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Carla Luiza da Silva Ávila.

Ethics declarations

Conflicts of interest

The authors declare that they have no conflict of interest.

Research involving human participants and/or animals

Not applicable

Informed consent

Not applicable

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 distributed under the terms of the Creative Commons Attribution 4.0 International License (, which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Santos, A.d.O.d., Ávila, C.L.d.S., Soares, C. et al. Lactic acid bacteria diversity in corn silage produced in Minas Gerais (Brazil). Ann Microbiol 69, 1445–1459 (2019).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: