Effects of storage temperature and ensiling period on fermentation products, aerobic stability and microbial communities of total mixed ration silage

Authors


Correspondence

Naoki Nishino, Department of Animal Science, Graduate School of Natural Science and Technology, Okayama University, Okayama 700-8530, Japan. E-mail: j1oufeed@cc.okayama-u.ac.jp

Abstract

Aims

To examine how storage temperatures influence ensiling fermentation, aerobic stability and microbial communities of total mixed ration (TMR) silage.

Methods and Results

Laboratory-scale silos were stored at 5, 15, 25 and 35°C for 10, 30 and 90 days. If silage was stored at 5°C, fermentation was weak until day 30, but acceptable lactic acid production was observed on day 90. The ethanol content was higher than the acetic acid content when stored at 15 and 25°C, whereas the ethanol content was lower when stored at 35 than at 25°C. Aerobic deterioration did not occur when silage was exposed to air at the same temperature at which it was stored. Although 10-day silages stored at 5 and 15°C deteriorated when the aerobic stability test was conducted at 25°C, heating was not observed in silages stored at 25 or 35°C or in any 90-day silages regardless of storage temperature. Denaturing gradient gel electrophoresis demonstrated that bands indicative of Lactobacillus plantarum and Lactobacillus delbrueckii were less prominent, while bands indicative of Lactobacillus panis became more distinct in silages stored at high temperatures. Bands of Kluyveromyces marxianus were seen exclusively in silages that were spoiled at 25°C.

Conclusion

High ambient temperature enhances acetic acid production in TMR silage. Lactobacillus panis may be associated with changes in the fermentation products due to differences in storage temperature.

Significance and Impact of the Study

The role of Lacto. panis in ensiling fermentation and aerobic stability is worth examining.

Introduction

The production of total mixed ration (TMR) silage has been employed in Japan to preserve and utilize wet by-products as ruminant feed. Food and beverage by-products often have high moisture contents; thus, blending with other feeds is necessary prior to use, even if good preservation can be achieved by ensiling. Wet brewers' grains and soybean curd residue are often used as the main ingredients and are mixed with lower proportions of dry feeds, such as grass hay, legume hay, maize grain, wheat bran and beet pulp. The levels of dry matter (DM), crude protein and total digestible nutrients are usually 500–600 g kg−1, 160–180 g kg−1 DM and 720–740 g kg−1 DM, respectively, and the product is usually commercially available in the form of a portable (300–400 kg) bag silo. Most manufacturers do not have enough space to store large quantities of silage; hence, a short ensiling period is preferable to enable shipment within 1 month of production.

An interesting property of TMR silage is its high aerobic stability after silo opening; spoilage does not take place for as long as 7 days, even in the summer, although nonensiled TMR easily deteriorates within 1–2 days (Nishino et al. 2004). Resistance to aerobic spoilage was observed even when more than 106 colony forming units (CFU) g−1 of yeasts were counted at silo opening (Nishino et al. 2004); typically, silages with over 105 CFU g−1 of yeasts are prone to spoil when exposed to air (McDonald et al. 1991). In addition, when ensiling is prolonged for several months, yeast counts fall below detectable levels (Nishino et al. 2004), which further improves the silage stability due to a lack of undesirable micro-organisms. From such aerobically stable silage, we previously demonstrated that Lactobacillus buchneri is the predominant species amongst lactic acid bacteria (LAB), and we confirmed the ability of Lact. buchneri to inhibit aerobic spoilage in grass and maize silages (Nishino et al. 2003; Nishino and Touno 2005). These results indicated that Lact. buchneri might be involved in silage stability; however, resistance to spoilage was seen in silages without detectable Lact. buchneri (Wang and Nishino 2008). High aerobic stability was also seen in a study from Israel, in which large-scale (700–800 kg) TMR silage was prepared using round bales wrapped in stretch polyethylene film (Weinberg et al. 2011); therefore, high shelf life can be expected, regardless of the ingredient composition and manufacturer location.

Unlike usual crop silage, TMR silage can be produced in any season. Because short-term storage is preferable, the storage temperature during the ensiling process may remain low in winter products and high in summer products. In our survey of commercial products, the acetic acid content in TMR silage from warm conditions was higher than that in TMR silage from cold conditions, whereas the opposite trend was observed for ethanol content (Wang and Nishino 2010). Many studies have reported that hot conditions (35–40°C) can lower the lactic acid production and aerobic stability (Yokota et al. 1991; Nishino and Uchida 1998; Weinberg et al. 1998; Ashbell et al. 2002). Seasonal changes in the fermentation products can be attributed to differences in the storage temperature; however, few studies have examined how cold storage may affect ensiling fermentation and aerobic stability. Moreover, seasonal changes in the fermentation might not be solely attributable to changes in storage temperature because diurnal and day-to-day variations also exist. Therefore, the effect of storage temperature is worth examining.

In this study, TMR silage was prepared in laboratory-scale silos and stored at 5, 15, 25 and 35°C to examine how storage temperature influences ensiling fermentation and aerobic stability. The aerobic deterioration test was performed both at the storage temperature and at 25°C because silage stored in cold conditions might deteriorate in warm conditions. Summer products of commercial TMR silage, produced using the same recipe as that used for the laboratory-scale silage, were also examined to determine how data for laboratory-scale silos related to those for practical-scale silos.

Materials and methods

Ensiling

A pre-ensiled TMR mixture, which was formulated to prepare commercial TMR silage, was obtained from a feed company on 25th July 2010. Approximately 30 ingredients were used to prepare the mixture; the main dried ingredients were cracked maize, rapeseed meal, sugar beet pulp and sudangrass hay, and the main wet ingredients were brewer's grains and soybean curd residue. Three hundred grams was placed in a plastic pouch (Hiryu BN12; Asahi Kasei Pax, Tokyo, Japan) and then tightly packed using a commercial vacuum sealer (SQ-303; Asahi Kasei Pax). The size, thickness and oxygen permeability of the pouch were 270 × 400 mm, 0·075 mm and 44 ml m−2 atm−1 per day, respectively. Ensiling mixtures were made in triplicate; the silos were stored at 5 (refrigerator), 15 (incubator), 25 (air-conditioned room) and 35°C (incubator) for 10, 30 and 90 days.

Commercial TMR silage was produced with the same recipe that was used for preparing the laboratory-scale silage and stored outside in a transportable bag (300 kg); samples of this commercial TMR silage were taken on the same day when laboratory silage was prepared. First, five grab samples were removed from the upper outside, upper inside, centre, bottom outside and bottom inside locations, and then, these samples were mixed well to make a composite. One representative sample was taken from the composite, and this procedure was repeated for another commercial TMR silage, which was produced on the same day by the feed company. Our preliminary study indicated that the fermentation product content and bacterial community did not vary with the sampling location (data not shown). Duration of storage was 35 days at the time of sampling for these commercial TMR silages.

Aerobic stability test

After the silage was opened, 100 g of the contents was put into two polyethylene bottles (500 ml) without compaction. The top of the bottle was kept uncovered and exposed to air for 7 days at 25°C or at the storage temperature for ensiling. Silage temperature was monitored every hour using an electronic thermal recorder. The silage was considered to have deteriorated when the temperature was 2°C more than the ambient temperature.

Chemical analyses and microbial enumeration

Silage samples were taken at the time of opening the silo and after conducting the aerobic stability test. Dry matter contents were determined by drying the material in an oven at 60°C for 48 h. Silage pH, lactic acid content, short-chain fatty acid content and alcohol content were determined from water extracts. Lactic acid, acetic acid and ethanol contents were determined by an ion-exclusion polymeric HPLC method with refractive index detection (Wang and Nishino 2010).

Bacterial counts were estimated for LAB using de Man, Rogosa and Sharpe agar (CM0361; Oxoid Ltd., Hampshire, UK) and for enterobacteria (ENB) using violet red bile agar (CM0107; Oxoid Ltd.). Yeasts and moulds were enumerated on spread plates of potato dextrose agar (CM0139; Oxoid Ltd.) that had been acidified to pH 3·5 with sterilized lactic acid. The plate cultures were incubated at 30°C for 3 days. Water-soluble carbohydrates (WSC) were extracted with 800 ml l−1 (v/v) ethanol, and the contents were determined by the phenol-sulfuric acid method (Parvin et al. 2010).

Denaturing gradient gel electrophoresis

Bacterial and fungal DNA were purified using a commercial kit (DNeasy Tissue kit; Qiagen, Germantown, MD, USA) according to the manufacturer's instructions. PCR was used to amplify a variable (V3) region of the bacterial 16S rRNA gene (Nishino et al. 2012), using the forward primer GC357f (5′-CGCCCGCCGCGCGCGGCGGGCGGGGCGGGGGCACGGGGGGCCTACGGGAGGCAGCAG-3′) and the reverse primer 517r (5′-ATTACCGCGGCTGCTGG-3′). For the amplification of the fungal 18S rRNA gene, the forward primer NS3 (5′-CGCCCGCCGCGCGCGGCGGGCGGGGCGGGGGCACGGGGGGGCAAGTCTGGTGCCAGCAGCC-3′) and the reverse primer YM951r (5′-TTGGCAAATGCTTTCGC-3′) were used (Li and Nishino 2011). The PCR protocol for bacterial denaturing gradient gel electrophoresis (DGGE) was as follows: initial denaturation at 95°C for 10 min; 30 cycles of denaturation at 93°C for 30 s; annealing at 65°C (first 10 cycles), 60°C (second 10 cycles) and 55°C (last 10 cycles) for 30 s; and extension at 72°C for 1 min, followed by a final extension at 72°C for 5 min. The PCR protocol for fungal DGGE was as follows: initial denaturation at 95°C for 10 min; 30 cycles of denaturation at 93°C for 1 min; annealing at 55°C (first 10 cycles), 50°C (second 10 cycles), and 45°C (last 10 cycles) for 1 min; and extension at 72°C for 3 min, followed by a final extension at 72°C for 5 min. PCRs were carried out in a PCR thermal cycler (Dice; Takara Bio Inc., Shiga, Japan).

The GC-clamp PCR products were separated according to their sequences with a DCode Universal Mutation Detection System (Bio-Rad Ltd., Tokyo, Japan). The samples were applied directly onto 100 g l−1 (w/v) polyacrylamide gels prepared with a denaturing gradient of 25−50% for bacteria and 10−35% for fungi using urea and formamide, respectively (7 mol l−1 urea and 400 ml l−1 formamide as 100% denaturants). Electrophoresis was performed at a constant voltage of 150 V for 12 h at 60°C. After electrophoresis, the gels were stained with SYBR Green (Cambrex Bio Science Inc., Rockland, ME, USA) and photographed under UV illumination.

Cloning and sequencing of DGGE bands

Selected bands were excised from the DGGE gels, and each band was placed in 10 μl of sterilized water at 4°C overnight to allow diffusion of the DNA. Extracted DNA was amplified by PCR using the 357f (without GC-clamp) and 517r primers for bacteria and the NS3 (without GC-clamp) and YM951r primers for fungi. The PCR products were purified using a commercially available kit (GeneClean Kit; Qbiogene, Carlsbad, CA, USA). The purified PCR products were cloned into the pTAC-1 vector, and the resulting plasmids were transformed into Escherichia coli strain DH5α competent cells (DynaExpress TA cloning kit; BioDynamics Laboratory Inc., Tokyo, Japan). Colony PCR was performed using primers for the vector (M13f and M13r) to confirm the presence of the correct inserts, and 1 colony per DGGE band was randomly selected for subsequent species identification.

The sequencing reactions were carried out using a BigDye® Terminator v3·1 Cycle Sequencing Kit (Applied Biosystems Inc., Foster City, CA, USA), and DNA base sequences were analysed using an ABI PRISM® 310 sequencer (Applied Biosystems Inc.). Searches of the GenBank database were performed with the BLAST program to determine the closest relatives of partial 16S rRNA (bacteria) and 18S rRNA (fungi) gene sequences. Sequences were considered positively identified when more than 99% of the sequence was identical to a sequence in the BLAST database.

Statistical analysis

Data were subjected to a two-way analysis of variance, with storage temperature and storage periods as the main variables. Because significant interaction was seen for many variables, the effect of storage temperature was examined for each storage period by a one-way analysis of variance, followed by a multiple comparison by Tukey's test. These analyses were carried out using JMP software (version 7; SAS Institute, Tokyo, Japan).

Results

Small amounts of lactic acid, acetic acid and ethanol were found in the pre-ensiled TMR mixture and LAB, yeasts and ENB were detected at 107, 106 and 105 CFU g−1, respectively (Table 1).

Table 1. Chemical and microbial composition of pre-ensiled total mixed ration mixture used for preparing the laboratory-scale silage
 Pre-ensiled material
  1. Values are means of duplicate analysis.

Dry matter (g kg−1)566
Crude protein (g kg−1DM)15·5
Water-soluble carbohydrate (g kg−1DM)72·3
pH5·73
Lactic acid (g kg−1DM)9·54
Acetic acid (g kg−1DM)1·61
Ethanol (g kg−1DM)2·16
Lactic acid bacteria (log cfu g−1)7·66
Yeasts (log cfu g−1)6·61
Enterobacteria (log cfu g−1)5·70

Differences in storage temperature resulted in marked differences in fermentation products on day 10 (Table 2). Fermentation occurred marginally in TMR silage stored at 5°C, whereas acceptable levels of lactic acid production were detected in TMR silage stored at temperatures higher than 15°C. The lactic and acetic acid contents increased as the storage temperature increased, whereas the ethanol content was higher than the acetic acid content when stored at 15 and 25°C. Ethanol content was lower in silage stored at 35°C than in silage stored at 25°C. 1-Propanol content was less than 5·0 g kg−1 DM in TMR silage stored at lower than 25°C; however, when stored at 35°C, the content increased to 13·8 g kg−1 DM, which was higher than the ethanol content. LAB were counted at 108–9 CFU g−1 in silage stored at 5, 15 and 25°C, but the number decreased to about 106 CFU g−1 when stored at 35°C. Although high quantities of yeasts (106–107 CFU g−1) were found in TMR silages stored at 5 or 15°C, the yeast content was below detectable levels when stored at 25 or 35°C. ENB counts were 104 CFUg−1 only in TMR silage stored at 5°C.

Table 2. Chemical and microbial composition of laboratory-scale total mixed ration silage stored at various temperatures for 10, 30, and 90 days
 10 daysSE30 daysSE90 daysSE anova
5°C15°C25°C35°C5°C15°C25°C35°C5°C15°C25°C35°CPTP×T
  1. P, storage period; T, storage temperature; P × T, interaction between storage period and storage temperature; NS, not significant.

  2. Means of triplicate silages. Values in the same storage period with different superscript letters (a–d, e–h, i–k) are significantly different (P < 0·05).

  3. *P < 0·05; **P < 0·01.

Dry matter (g kg−1)5405175305510·90539e,f524f543e,f562e0·77545j528j536j612i0·72 ** ** *
pH5·25a4·61b4·40c4·30d0·024·83e4·46f4·38f,g4·29g0·024·31i4·27j,k4·28i,j4·22k0·01 ** ** **
Lactic acid (g kg−1DM)13·5c44·9b54·3a55·4a1·0028·6g52·2f59·7ef61·5e1·9543·4j56·8i57·3i50·0i2·78 ** ** **
Acetic acid (g kg−1DM)2·21d7·53c12·3b17·2a0·287·69h11·3g16·1f19·1e0·5910·1j15·2i17·2i16·1i0·86 ** ** **
Ethanol (g kg−1DM)3·09d26·1a20·8b8·10c0·837·33g25·1e18·5f6·41g1·1913·9j22·6i11·4j3·90k1·46NS ** **
1-Propanol (g kg−1DM)0·00c0·36c3·35b13·8a0·470·00g3·30f5·15f11·7e0·451·88k2·81i,j2·63i,j3·40i0·29 ** ** **
Lactic acid bacteria (log cfu g−1)8·26b9·38a8·70b5·93c0·148·96e9·15e7·44f5·63g0·058·76i7·67i5·04j3·82k0·02 ** ** **
Yeasts (log cfu g−1)6·857·16<2·00<2·006·60<2·00<2·00<2·00<2·00<2·00<2·00<2·00
Enterobacteria (log cfu g−1)4·54<2·00<2·00<2·00<2·00<2·00<2·00<2·00<2·00<2·00<2·00<2·00

Storage temperature also affected fermentation products on day 30. Weak fermentation appeared to have continued in TMR silage stored at 5°C. From day 10, increases in the lactic acid content were observed in TMR silage stored at 15°C. There were no differences in the pH value or lactic acid content between the silage stored at 15°C and at 25°C. Similar to the results on day 10, the lactic acid, acetic acid and 1-propanol contents on day 30 increased as the storage temperature increased. Ethanol was the second most common product in TMR silage stored at 15°C; however, the content did not increase between days 10 and 30, whereas the lactic and acetic acid contents did increase. High numbers of LAB (about 109 CFU g−1) were counted in TMR silages stored at 5 and 15°C, and the LAB population decreased as the storage temperature increased. Although yeasts were found at 106 CFU g−1 when silage was stored at 5°C, the numbers were below detectable levels if stored at the higher temperatures. On day 30, ENB were not detected in TMR silage regardless of storage temperature.

After storing for 90 days, the pH decreased to 4·31 even in TMR silage stored at 5°C, and the lactic acid, acetic acid and ethanol contents were 43·4, 10·1 and 13·9 g kg−1 DM, respectively. Few changes were seen in the contents of the fermentation products between days 30 and 90 in silages stored at greater than 15°C. In TMR silage stored at 25°C, more acetic acid than ethanol was observed on day 90; however, the ethanol content was higher than the acetic acid content in TMR silage stored at 15°C. The LAB counts were high (about 108 CFU g−1) in TMR silages stored at 5 and 15°C, but the population decreased to 103 CFU g−1 in silages stored at 35°C. No yeasts or ENB were detected on day 90 in any TMR silages, regardless of storage temperature.

The lactic acid, ethanol and 1-propanol contents of commercial TMR silage were similar to those of laboratory-scale TMR silage stored at 35°C for 30 days (Table 3). However, intensive acetic acid production was seen in commercial TMR silage; the content (41·6 g kg−1 DM) was more than double the content (19·1 g kg−1 DM) of laboratory-scale TMR silage.

Table 3. Chemical composition of commercial total mixed ration silage prepared using a different batch of the same ingredient mixture used for preparing the laboratory-scale silage
 Commercial total mixed ration silage
  1. Values are means of two bag silages. Live counts were not determined for these silages.

Dry matter (g kg−1)570 ± 6·36
pH4·14 ± 0·00
Lactic acid (g kg−1DM)63·9 ± 9·90
Acetic acid (g kg−1DM)41·6 ± 6·58
Ethanol (g kg−1DM)9·38 ± 3·43
1-Propanol (g kg−1DM)6·43 ± 2·11

Changes in silage temperature and fermentation product content during the 7-day aerobic stability test demonstrated that no aerobic deterioration occurred when the TMR silage was exposed to air at the same temperature at which it was stored (data not shown). However, when the aerobic stability test was conducted at 25°C, 10-day TMR silages stored at 5 and 15°C and 30-day TMR silage stored at 5°C deteriorated with a significant increase in silage temperature observed within 48 h after silo opening. After 90 days of storage, heating was not observed in any of the TMR silages, regardless of the storage temperature and even if the aerobic stability test was performed at 25°C.

The patterns of DGGE bands were the same on day 10 in both pre-ensiled TMR mixtures and TMR silages stored at 5°C (Fig. 1). Bands indicative of Lact. helveticus (band 1), Lact. brevis (band 2), Lactobacillus sp. (band 6), Tetragenococcus halophilus (band 7) and Lact. buchneri (band 8) were distinct in the 10-day TMR silage stored at 5°C. In addition to these bacteria, Lact. plantarum (band 4) appeared in TMR silage stored at 15°C. When TMR silage was stored at 25°C, bands for Lactobacillus sp. (the same position as band 3), Lact. panis (band 9) and Lactobacillus sp. (band 10) were detected from day 10. In silage stored at 35°C, bands indicative of Lact. plantarum and Lact. delbrueckii (band 13) were faint, and bands of Lact. panis (bands 11) and Lact. pontis (band 14) appeared from the start of ensiling. When ensiling was prolonged to 30 and 90 days, Lact. plantarum became detectable only in TMR silage stored at 5°C. In addition, longer ensiling led to bands indicative of Acinetobacter baumannii (band 5) in TMR silage stored at 5 and 15°C. The DGGE band patterns of commercial TMR silage resembled those of laboratory-scale TMR silage stored at 35°C; bands indicative of Lactobacillus sp. (the same position as band 3), Lact. panis (the same position as band 11) and Lact. pontis (the same position as band 14) were commonly found in these TMR silages.

Figure 1.

Bacterial communities of laboratory-scale and commercial total mixed ration silages prepared from the same pre-ensiled material. Laboratory-scale silages were stored at various ambient temperatures for 10, 30 and 90 days, while commercial silages were stored outside in the summer for about 1 month.

Although the plate culture detected yeasts only in the 10-day TMR silages stored at 5 and 15°C and 30-day TMR silage stored at 5°C, the DGGE analysis detected yeasts in all TMR silages (Fig. 2). Bands indicative of Pichia kudriavzevii (band 20) were observed regardless of storage temperature and storage period. Bands for Kluyveromyces marxianus (band 18) were seen in the pre-ensiled TMR mixture and TMR silages stored at 5°C over the course of ensiling. Distinct bands of Candida milleri (band 19) were found in the 30-day and 90-day TMR silages stored at 15 and 25°C. The DGGE band patterns of the commercial TMR silage resembled those of the laboratory-scale TMR silage, with distinct bands for C. milleri (the same position as band 19) and P. kudriavzevii (the same position as band 20).

Figure 2.

Fungal communities of laboratory-scale and commercial total mixed ration silages prepared from the same pre-ensiled material. Laboratory-scale silages were stored at various ambient temperatures for 10, 30 and 90 days, while commercial silages were stored outside in the summer for about 1 month.

Discussion

Intensive ethanol production was seen exclusively in the initial ensiling in TMR silages stored at 15 and 25°C, while weak production continued throughout the ensiling process in those stored at 5°C. Prominent bands of C. milleri observed for TMR silage stored at 15 and 25°C coincided with intensive ethanol production at these temperatures, whereas the plate culture did not detect yeasts in the 10-day TMR silage stored at 25°C. The growth of C. milleri in liquid culture has been shown to terminate at 36°C (Gänzle et al. 1998), and indeed, bands for C. milleri were faint or invisible in TMR silage stored at 35°C. Therefore, activation and inactivation of C. milleri might account for differences in ethanol production in TMR silage stored at different temperatures. Likewise, our results appeared to agree with the ‘baker's rule’ that compared with high temperatures, low temperatures during sourdough fermentation (20–26°C) are better for yeast growth (Gänzle et al. 1998).

Enhanced acetic acid production at high storage temperature (40°C) was demonstrated with guinea grass and maize silages (Kim and Adesogan 2006; Liu et al. 2011); however, the opposite results have also been reported with wheat and maize silages (Ashbell et al. 2002). In our experiments, Lact. panis was observed in silages stored at 25 or 35°C, whereas Lact. plantarum was faint or undetectable in silages stored at 35°C. Our previous survey also indicated that, even though Lact. plantarum is found in commercial TMR silage produced in the winter, spring and autumn, the LAB species is eliminated, and Lact. panis is present in silage produced in the summer (Wang and Nishino 2010). These changes in the bacterial composition could account for the increases in the acetic acid content observed at higher temperatures; however, it is difficult to explain how the hetero-fermentative LAB is activated with increases in storage temperature. Moreover, it is not known whether similar changes in the bacterial community may occur in conventional crop silages. Further research is required to clarify the effect of storage temperature on the activity of silage micro-organisms; however, increases in the acetic acid content in the higher temperature conditions can help suppress fungal activity in silage after exposure to air.

Aerobic deterioration did not occur in the 10-day TMR silages stored at 25 and 35°C, which showed undetectable (<102 CFU g−1) levels of yeast populations at silo opening. When the aerobic stability test was carried out at 25°C, heating was observed in the 10-day TMR silages stored at 5 and 15°C and in 30-day TMR silage stored at 5°C. In those silages, more than 106 CFU g−1 of yeasts were found at silo opening, but the yeast populations became undetectable when storage was prolonged, and these silages showed resistance to spoilage. These findings indicate that aerobic spoilage could take place if TMR silage is prepared at low temperatures, opened after a short storage and then transferred to a warm environment before feeding to animals. Our results suggest that spoilage should not occur after long-term storage regardless of the storage temperature or if the TMR silage is used in a region where the climatic conditions are similar to the silage-producing factory regardless of the storage period. If aerobic spoilage occurs, it can be attributed to the activity of yeasts. In this study, TMR silages became resistant to deterioration if the acetic acid content exceeded c. 10 g kg−1 DM and the yeast population decreased to below detectable levels.

Similar to our previous findings (Wang and Nishino 2010), Lact. panis was found in laboratory silages stored at 25 and 35°C and in commercial silages produced in the summer; although both the pre-ensiled material and the commercial product were obtained from a feed company, we did not examine the products in our previous study. Therefore, the association of Lact. panis with ensiling at higher temperature may be a common feature in TMR silage regardless of the region and factory. Lactobacillus panis is a member of the sourdough LAB, and it grows at 30–45°C, but not at 15°C (Wiese et al. 1996). Isolation of Lact. panis from the TMR silage and use of this isolate as an inoculant in silage are worth examining. However, we have not succeeded in obtaining culturable forms of LAB species until now.

Our previous studies have shown that Lact. buchneri is present over the course of fermentation and confers resistance to spoilage in TMR silage (Wang and Nishino 2009). Because Lact. buchneri was found in many samples in that study, the high stability of TMR silage could be attributed to the activity of this LAB species. In addition, the large amount of 1-propanol in 10-day TMR silage stored at 35°C suggested that during the first 10 days of ensiling, Lact. buchneri produced 1,2-propanediol and then Lact. diolivorans metabolized the diol to 1-propanol (Krooneman et al. 2002). However, the DGGE band indicative of Lact. buchneri was observed even in the pre-ensiled material and in TMR silages that subsequently spoiled in the warm environment. Moreover, the typical changes in fermentation by Lact. buchneri treatment, namely reduction in lactic acid and increase in acetic acid upon prolonged storage were not observed in the present study. The DGGE analysis can give us qualitative or semi-quantitative information, but it does not demonstrate how the detected species may function in the samples. Therefore, even though Lact. buchneri was detectable across the TMR silages regardless of storage temperature and storage period, the ability to resist spoilage needs to be evaluated by comparing the data on bacterial community and fermentation products. It is not known whether the metabolism of Lact. buchneri is influenced by the ambient temperature during ensiling.

We observed few variations in the fungal community of the TMR silage due to changes in temperature. Species detected by DGGE in this study were previously found in silage or other fermented foods; C. milleri was isolated from aerobically deteriorated maize silage (Middelhoven and Van Baalen 1988), and K. marxianus and K. exigua were found in fermented dairy products (Zhou et al. 2009). Bands of K. marxianus were seen exclusively in the TMR silages that were spoiled, including the silages that were stored at 5°C for 10 and 30 days before the aerobic stability test was conducted at 25°C. Therefore, although reduction of yeast populations may be the primary reason why TMR silage became resistant to spoilage over the course of fermentation, elimination of K. marxianus could also be a factor. The detection of C. milleri might further suggest that TMR silage and sourdough resemble each other with regard to microbial community and long shelf life in the presence air; however, cultivable yeasts were not found in long-term storages of silage and aerobically stable TMR silages. The roles of yeasts in the long shelf life are probably much smaller in TMR silage than in sourdough.

Based on both fermentation products and microbial communities, commercial TMR silage produced in the summer was similar to laboratory-scale TMR silage stored at 35°C for 30 days, indicating that our plastic pouch method should serve as a good model to examine the ensiling process on a more practical scale. It is difficult to explain why commercial silage had more than double the acetic acid content than laboratory-scale silages; however, commercial products were left outside in the summer, and thus, the silage temperature may have been greater than 35°C for some period. Silage temperature should have levelled off at 35°C if a small amount was stored in the incubator. Therefore, the outdoor storage of commercial products could intensify acetic acid production at higher temperatures.

In conclusion, high ambient temperature enhances the production of acetic acid during ensiling and sourdough LAB, particularly Lact. panis, may be associated with changes in fermentation products and the aerobic stability of TMR silage.

Acknowledgements

A part of this study was supported by the Ministry of Agriculture, Forestry and Fishery of Japan (Research for production of valuable livestock by feeding self-sufficient forage crops) and by a Grant-in-Aid for Scientific Research (No. 24580390) from the Japan Society for the Promotion of Science.

Ancillary