1. Introduction
Red clover (
Trifolium pratense L.) and lucerne (
Medicago sativa L.) are important legume forages with high nutritional value and are difficult to ensile, especially at low DM contents, because of the low content of water-soluble carbohydrates (WSC) and high crude protein (CP) content and buffering capacity (BC) [
1]. Inoculation with homofermentative lactic acid bacteria (LAB) can improve fermentation quality by accelerating and increasing the production of lactic acid (LA), thus achieving acidic conditions more rapidly [
2,
3,
4]. However, inoculants may not effectively improve forage preservation if the availability of fermentation substrates are insufficient to permit LAB growth [
5]. Pre-fermented juice (PFJ) is a fermentation juice made from fresh forage containing natural epiphytic microorganisms, which can subsequently be used as a silage inoculant. The addition of PFJ at ensiling was more effective at improving the fermentation quality than LAB inoculation in some previous studies [
2] because epiphytic LAB from PFJ stimulates LA production more effectively, especially in high-moisture legume silages, which are typically difficult to ensile [
6,
7].
Several authors have reported improvements in fermentation quality of high-moisture legume silages by adding PFJ [
2,
8]. Denek et al. [
9] reported that adding PFJ prepared from barley, wheat and grass herbages improved the fermentation quality of lucerne silage. Similarly, PFJ from lucerne was regarded as a good source of LAB for ensiling lucerne, inhibiting proteolysis and production of non-protein nitrogen (NPN) [
2]. Ohshima et al. [
10] showed that PFJ prepared from ensiling material is more effective at improving the fermentation quality than that prepared from other forage sources, possibly due to the better survival and growth of epiphytic LAB during ensiling. However, to our knowledge, no studies have reported the difference in composition of lactic acid bacteria in PFJs made from various forages, which is critical to explain the mode of action of PFJ and can be useful to develop more effective inoculants for various forages.
Two PFJ preparation methods, chopping and chopping + blender maceration, are used in the silage processing industry. While chopping reduces forage particle length, chopping + blender macerating samples reduce it even further. The differences between these methods in the LAB count and quality of PFJ have not been studied previously. The first objective of this study was to evaluate the effects of chopping and chopping + blender maceration on the characteristics of PFJ prepared from red clover and lucerne. The second objective was to identify the LAB species isolated from the PFJs and determine fermentation quality of red clover silage and lucerne silage inoculated with PFJ made from lucerne or red clover. We hypothesized that the chopped PFJ would have better quality compared with the chopping + blender macerating PFJ. Additionally, we hypothesized that PFJ from the forage being ensiled will be more effective at improving fermentation than that from the alternative forage.
4. Discussion
In the current study, the glucose was added to the PFJ to act as a substrate that would promote a good fermentation in the presence of an adequate number of LAB. Supplementing WSC to PFJ for facilitating fermentation was previously suggested, and glucose was regarded as the best sugar substrate for the preservation of silage [
5,
10]. The higher LAB count of PFJs prepared from chopping than chopping + blender maceration indicated that chopping is a better processing method for preparing PFJ (
Table 1). Consequently, the PFJ-RC and PFJ-LC were selected as silage additives in the present study. The LAB counts in the PFJs (7.57 and 7.80 log CFU/mL in PFJ-RC and PFJ-LC, respectively) were slightly lower than those reported by Ohshima et al. [
6,
7,
10]. However, spraying 3.0 mL/kg (FW) of PFJ on materials in the present study were still adequate to achieve the LAB count (10
5 CFU/g FW) required for an adequate fermentation [
1].
In the study of Tao et al. [
23], only two genera of LAB were reported in PFJ prepared from lucerne. However, in the present study, seven and eight genera were isolated and identified from PFJ-RC and PFJ-LC, respectively. The differences between both studies might be attributed to different laboratory methods used to identify LAB. While Tao et al. [
23] used selective agar media-based culturing, 16S rRNA gene sequencing, which has better coverage, was used in the present study. Six common LAB species were identified in PFJ-RC and PFJ-LC, probably because red clover and lucerne were harvested from the same experimental farm. Previous studies have detected a variety of LAB species in silages including
L. plantarum,
P. pentosaceus,
L. brevis,
L. mesenteroides and
L. lactis [
24,
25,
26]. In our study, all of these species were isolated from both PFJ-RC and PFJ-LC and accounted for over 60% of all identified LAB species. In addition,
L. paracasei was identified only in PFJ-RC, while
W. cibaria and
P. acidilactici were only found in PFJ-LC (
Table 2). The wide range of LAB species present in PFJs and high LAB count indicated the suitability of using PFJs as silage inoculants to facilitate fermentation. In addition, the different LAB composition of PFJ-RC and PFJ-LC indicated that red clover and lucerne contain different epiphytic LAB species.
Due to the low LAB count and WSC content and the higher buffering capacity of red clover and lucerne before ensiling (
Table 3), it is difficult to achieve an efficient fermentation during ensiling [
21]. This was reflected by the low lactic acid concentration and high pH and NH
3-N/TN in Control silages in the present study (
Table 4). The propionic and butyric acids were detected only in Control silages at a very low level (
Table 4) because the high DM content (401 and 398 g/kg, respectively) and ideal anaerobic environment may have caused a certain decrease in activity of undesired microorganisms during ensiling process in Control silages. The ratios of lactic acid to acetic acid in Control silages were below one (
Table 4) and indicated that abnormal fermentations had occurred in the silages [
27,
28], resulting from activity of
Enterobacteriaceae dominating the bacterial community in Control silages during fermentation process.
Enterobacteriaceae could thrive in anaerobic and weak acidic conditions [
29] and ferment WSC and lactic acid to acetic acid, succinic acids, ethanol, or 2,3-butanediol [
30,
31]. The presence of
Enterobacteriaceae may have accounted for the low levels of WSC (20.5 and 20.5 g/kg DM, respectively), the low lactic acid (7.56 and 8.58 g/kg DM, respectively) and the high pH (5.31 and 5.43, respectively) in Controls of both silages (
Table 4). Based on these above mentioned results, it is necessary to ensile red clover and lucerne with inoculants. However, as the valeric acid, caproic acid, ethanol, propanol, 1,2-propandiol, and DM losses were not analyzed in the present study, the effect of inoculating PFJ on fermentation quality of silages could not be evaluated accurately. Thus, additional analyses of the alcohols, acids, and DM losses stated above would have helped in better understand the effect of PFJ; this needs further study to fully evaluate the influence of ensiling red clover and lucerne with PFJ.
Successful silage fermentation mainly depends on the ensiling technique, properties of the inoculant, characteristics of plants ensiled, epiphytic microflora, and climatic conditions [
32]. Ensiling legumes, which have a higher buffering capacity, with LAB can improve the fermentation quality and inhibit undesirable microorganisms [
33]. According to the evaluation system for fermentation quality of silages based on the contents of butyric and acetic acids [
34], the scores of all treatments in both silages were 100, and their mark was the highest (first) in the present study, because of the lower contents of butyric and acetic acids in all silage (<3.0 and <30 g/kg DM, respectively). The higher lactic-to-acetic ratio, LAB count, and lactic acid concentration and lower pH for PFJ treatments suggested homolactic fermentation dominated the ensiling process [
28]. Our results agreed with Filya et al. [
20] who showed that a number of commercial homolactic inoculants improved lucerne silage fermentation by shifting fermentation towards lactic acid and reducing the pH. In addition, Ohshima et al. [
10] reported that ensiling lucerne and Italian ryegrass with PFJ containing high concentrations (10
8 CFU/mL) of epiphytic LAB increased lactic acid concentration and accelerated the decrease in silage pH. Our results are also in agreement with other studies [
2,
9,
23], showing efficacy of PFJ at increasing LAB count and lactic acid and decreased pH, acetic acid, and NH
3-N/TN of silage (
Table 4). According to the typical suggested concentrations of common fermentation end products in legume silage [
28], in the present study, the PFJ treatments contained appropriate pH and lactic acid contents and had very low acetic acid in both silages, which resulted in a high lactic-to-acetic acid ratio in PFJ treatments.
The LAB of PFJ prepared from ensiling crops may grow better on the same forage during fermentation than the PFJ prepared from another crop [
10]. In the present study, the forages for preparing PFJ and ensiling were grown in the same field. Ensiling red clover with PFJ-RC reduced pH and NH
3-N/TN and increased LAB count, lactic acid concentration, and lactic-to-acetic acid ratio to a greater extent than PFJ-LC (
Table 4). Similarly, the PFJ-LC had more positive effect on the LAB, pH, lactic acid, and NH
3-N/TN of lucerne silages than PFJ-RC (
Table 4). The results suggested that inoculating lucerne or red clover with PFJ prepared from the ensiling material was more effective at improving fermentation parameters detected in the present study. This might be due to the materials prepared for PFJ and silage being collected from the same field, and the LABs in PFJ are more familiar with the physico-chemical properties of the silage [
35]. Similarly, Ali et al. [
36] revealed that the microbiota from red clover had a better effect on bacterial community succession and fermentation quality in red clover silage than that from maize and sorghum. In addition, Sun et al. [
35] also found that the LAB from whole-plant corn silage had a better promotion effect on microbial succession and fermentation of whole-plant corn silage than those from
Elymus sibiricus silage.
The NH
3-N is part of the non-protein in silage and NH
3-N/TN indicates the degree of silage preservation during fermentation [
37,
38]. In the present study, although the Control silages had the suggested contents of NH
3-N/TN according to Kung et al. [
28], the PFJ treatments contained lower levels of NH
3-N/TN (
Table 3). This indicated that applying PFJs decreased NH
3-N/TN in both red clover silages and lucerne silages, because adding PFJ could effectively reduce activity of
Enterobacteriaceae.
Enterobacteriaceae are responsible for much of the NH
3-N formed from protein degradation and from the reduction of NO
3 [
32] in Control silages. This is in agreement with Wang et al. [
2] and Tao et al. [
23], who reported that lucerne silage treated with PFJ contained less NH
3-N/TN than the silage with commercial LAB additives in some cases. For both forages from the same experimental farm, the PFJ made from the ensiling material was more effective in well preserving silage than PFJ made from the other forage source.
Both PFJ treatments increased the
IVDMD of both silages, but PFJ-LC increased the
IVDMD to greater extent than PFJ-RC (
Table 5). The increased
IVDMD by ensiling red clover or lucerne with PFJs is consistent with Nishino and Uchida [
8], who reported that adding PFJ to lucerne at ensiling increased the
IVDMD. Similarly, Ellis et al. [
39] also reported that LAB inoculation can improve the digestibility of silage. The increased
IVDMD might be due to the satisfactory fermentation parameters detected in PFJ-treated silages contributing to the preservation of digestible nutrients in the present study, which were reflected by the
IVDMD correlating positively with lactic acid and lactic acid/acetic acid and negatively with pH, acetic acid, and NH
3-N/TN (
Table 6). Moreover, the plant cell wall negatively influences the digestibility of silage [
40], which was reflected by the negative correlation of
IVDMD with NDF and ADF for both silages in the present study (
Table 6). Similarly, previous studies also reported that the higher
IVDMD was detected in silages with lower NDF and ADF concentrations [
27,
41].