Livestock Research for Rural Development 20 (7) 2008 Guide for preparation of papers LRRD News

Citation of this paper

Nutritional value of soybean (Glycine max L. Merr.) silage fermented with molasses and inoculated with Lactobacillus brevis 3

C Tobía, E Villalobos*, A Rojas**, H Soto** and K J Moore***

 Department of Forages and Animal Nutrition, University Centroccidental Lisandro Alvarado, UCLA, Barquisimeto, Venezuela
carlostobia@ucla.edu.ve
*Grain and Seed Research Center (CIGRAS), University of Costa Rica
villalobos.enrique@gmail.com
**School of Animal Science, University of Costa Rica, San José, Costa Rica
augustor@cariari.ucr.ac.cr and hsoto@cariari.ucr.ac.cr
***Department of Agronomy, Iowa State University, USA
kjmoore@iastate.edu
 

Abstract 

Protein is often limiting in ruminant diets in the tropics. Soybean (Glycine max L. Merr.) varieties developed for grain production in tropical environments could potentially be harvested as forage to supplement protein in diets of ruminant livestock. The nutritional value of ‘CIGRAS-06’ soybean forage fermented with molasses and inoculated with Lactobacillus brevis 3 (Lb3) was evaluated. The above ground part of the soybean plants was harvested at the R6 (full seed) stage of development and chopped in approximately 2-cm sections. Molasses was applied in concentrations of 0, 3, 6 or 9% of forage fresh weight and the herbage was sprayed with an equivalent of 6.5 x 107 Lb3 colony forming units per ml of air, before storing in plastic micro-silos (1 kg).  

Molasses treatment resulted in a proportional increase in silage dry matter (DM), ash (As) and neutral detergent soluble carbohydrates (NDSC). Conversely, a dilution effect was observed in crude protein (CP), ether extract (EE) and neutral detergent fiber (NDF). Molasses improved silage sensorial indicators and reduced pH, NH3 as a percentage of total nitrogen (NH3/%TN), acetic (AA) and butyric (BA) acids concentration. Molasses increased lactic acid (LA) and energy concentration, with best results (72% of total digestible nutrients) obtained with the 3% treatment. Lb3 decreased silage NH3/%TN content and BA, but decreased EE concentration and DM. Although Lb3 effectiveness was not as obvious as that of molasses, a synergistic effect between both additives was observed for pH, AA, BA and NDF reduction.

Both additives improved the fermentation and production of high quality soybean silage that due to its high buffer capacity and relatively low soluble carbohydrate content, would otherwise be difficult to ensile.  

Keywords: soybean silage fermentation, nutritional value, molasses, Lactobacillus brevis 3, soybean forage preservation


Introduction

Protein is often limiting in ruminant diets in the tropics where tropical grasses are often the primary component (Rojas et al 1998). Although legume forage is of better quality than that of grasses (Weiss and Shockey 1991), its potential use in animal nutrition in the tropics is underexploited. 

Soybean (Glycine max L. Merr.) varieties developed for grain production in tropical environments could potentially be harvested as forage to supplement protein in diets of ruminant livestock. Soybean grain production in Costa Rica is limited to warm and flat areas with a seasonal rainfall pattern, where there is a possibility of harvesting a good quality seed during the dry season. However, soybean production requires a high investment and competes for resources with well established rice and sugar cane industries. Soybean grain production in the wet tropics is also negatively affected by high and continuous rainfall throughout the year that often limits soil preparation and planting and often impedes harvesting an acceptable product. Thus, imported soybean seed, after extracting the oil, is the main source of protein in balanced concentrate formulas for animal feed in Costa Rica. 

Soybean forage, when harvested at the R6 stage of development (Fehr and Caviness 1980) is comparable to that of alfalfa, both in quality and quantity (Hintz and Albretcht 1994), since a well established variety can produce between 1200 and 2000 kg/ha of protein. Additionally, the energy value of soybean forage is even higher than that of corn forage. At the R6 stage of development, soybean seeds have reached their highest dry matter content and represent the part of the plant with the best nutritional value (Hintz and Albretcht 1994).  

Soybean forage can be produced even in the wet tropics at a much lower cost than soybean grain, because it does not requires a dry season coincident with time of seed ripening nor any expensive harvesting machinery (Tobía and Villalobos 2004). A viable alternative to preserve and to take advantage of the nutritional quality of this forage is ensiling a well adapted soybean variety. In the past, it was not easy to find a soybean variety with good adaptation in the tropics. Most soybean genotypes introduced in tropical areas from Southern United States or Northern Brazil, flower too soon to reach an adequate plant development, due to photoperiodic flowering induction. However, the introduction of genes that delay flowering under short day conditions into local breeding programs has obviated the problem of soybean adaptation in these areas (Villalobos and Camacho 2003). The best seed producer has also been considered the best variety for forage in Northern Latitudes (Hintz and Albretcht 1994). However, during the last decade, forage soybean varieties  have been developed in different US institutions (Blount et al 2006).  In the tropics a soybean variety for forage may have a later maturity (2 or 3 weeks) than that selected for seed, provided that it is resistant to lodging, since the soybean forage is harvested one month earlier than the seed. Late maturity soybean genotypes in tropical conditions can be easily developed, transferring those genes that delay flowering under short-day conditions into local varieties (Villalobos and Camacho 2003). 

Soybean forage is often difficult to ensile due to its relatively high buffer capacity, its low soluble carbohydrate content (Bolsen et al 2001) and relatively low dry matter concentration (Mc Donald 1981). In order to solve this limitation, the use of lactic acid bacteria (LAB) and the addition of soluble carbohydrates have been recommended (Jaster 1995). That net effect of both additives is to increase lactic acid concentration, thus creating the appropriate environment for silage preservation. Muck and Kung (1997) recommend use of bacteria strains isolated from the forage to be ensiled, in order to achieve best results with the fermentation process. 

The objective of this research was to quantify those indicators associated with the fermentation process of soybean forage and to evaluate silage quality in response to the application of sugar cane molasses and LAB previously isolated from soybean micro silos and identified as Lactobacillus brevis 3 by Tobía et al (2003). 
 

Materials and methods 

Soybean forage 

The tropical soybean variety ‘CIGRAS-06’, developed at the Grain and Seed Research Center of the University of Costa Rica (Villalobos and Camacho 2003), was used. This variety was selected because of its high yield potential, lodging resistance, and disease and insect resistance when grown in different locations of Costa Rica and Nicaragua. The experimental plot (2.4 has) was located in Sarapiquí, Atlantic region of Costa Rica, under very wet conditions (4000 mm of rain per year) and adverse soil conditions (loam clay texture with a pH 4.8 and low fertility) (Tobía and Villalobos 2004). Plant population consisted of 192.000 plants per hectare, with a distribution of 16 plants per row meter and 80 cm between rows. Seed was inoculated with Bradyrhizobium japonicum immediately before sowing. Commercial fertilizer formula 10-30-10 was applied at a rate of 250 kg/ha at sowing time. Soybean plants were harvested at the R6 stage of development, which was identified when at least 1 pod in the 4 uppermost nodes in the main stem of the soybean plant has seeds completely full (Fehr and Caviness 1980). This stage of development, which is usually coincident with yellowing of lower leaves, has been determined to be optimum for forage (Hintz and Albrecht 1994).  

Treatments 

Treatments consisted of sugar cane molasses applied in concentrations of 0, 3, 6 or 9% of soybean forage fresh weight and fermentation with and without inoculation with Lactobacillus brevis 3 (Lb3) (Tobía et al 2003) at an approximate rate of 2.5 ml/kg of forage fresh weight, which is equivalent to 6.5 x 107 Lb3 colony forming units . Treated herbage (1 kg) was packed into plastic micro silos which were then evacuated, sealed, and stored for 45 days at 23 °C. 

Analysis of soybean forage 

The above ground part of the soybean plant was harvested and chopped in approximately 2-cm sections. Four samples were taken to determine dry matter concentration in an air flow oven at 60 C for 48 hours. Crude protein (CP) was determined using the methodology recommended by the AOAC (1990). Forage buffer capacity was estimated using methodology described by Playne and McDonald (1966). Neutral detergent soluble carbohydrates (NDSC) were determined using the procedure of Hall (1997). 

Sensory analysis of silage 

Sensory evaluation was conducted immediately after opening the micro silos, after 45 days in anaerobic conditions. The procedure used was that developed by Ojeda et al (1991) wich consists of assigning the highest value to the odor, followed by color and silage texture. The values assigned to these 3 parameters can add to a maximum of 100%. However, in this research, only odor and texture were evaluated, and consequently, the maximum score assigned to a good quality silage was 76%. Color was not taken in consideration, because molasses masks silage color.  

Chemical analysis of silage  

Two 20-g samples of soybean silage were taken from each micro silo. Each sample was stirred with 80 ml of distilled water for 4 hours in a mechanical shaker and filtered. The pH of the filtrate was obtained with a pH meter (Rojas 1985). The filtrate was centrifuged for 15 minutes at 2500 rpm. A 10-ml aliquot was taken from the supernatant, mixed with 10 ml of NaOH 40%, and distilled into a beaker containing 15 ml H3BO4 and then titrated with H2SO4. This quantity of H2SO4 was used to determine the amount of NH3 present in the sample (AOAC 1990).  

Another aliquot of the supernatant (3 ml) was drained through a cellulose acetate filter (Albetâ) with a pore diameter of 0.45 m. The filtrate was saved in glass vials and stored in a freezer at 5 C prior to organic acid analysis by gas chromatography. Acetic, butyric and lactic acids were separated and quantified using gas chromatography (Hewlett Packardâ model 5890 Series II) equipped with a HP-INNOWAX 19091N-133 column. The inlet temperature was 200 °C and the oven temperature was 160 °C. The carrier gas was helium at a flow rate of 1 ml/min.  

Ether extract (EE), dry matter concentration (DM) and ash content (As) of soybean silage were determined using the recommended procedures of the AOAC (1990). CP and NDSC were determined using similar procedures described by Hall (1997). Neutral detergent fiber (NDF), acid detergent fiber (ADF), hemi-cellulose, cellulose and lignin were determined using the procedure of Goering and Van Soest (1970). Nitrogen content in NDF and ADF was analyzed using the procedure recommended by the AOAC (1990). 

Total digestible nutrients (TDN) concentration was obtained by summing the energy from the following constituents: CP (ECP), fatty acid content in the EE (EFA), the NDF (ENDF) and the NDSC (ENDSC), minus a correction factor for fecal metabolic matter (CFMM) (Weiss 1999 and NRC 2001), as shown in the following equation:

%TDN = (ECP+ EFA + ENDF +E NDSC) - CFMM

where,

ECP = CP x e-1.2 . INAD
EFA = 0.94 x (EE-1) x 2.25
ENDF = 0.75 x [ (NDFCP – Lig) x (1- (Lig / NDFCP)).667]

ESCNDS = 0.98 x (100 – (NDFCP + CP +As +EE))

CFMM = Correction factor for fecal metabolic matter was fixed in 7, because the model is based on true digestibility and the TDN acts on the apparent digestibility.
INAD = Insoluble nitrogen in acid detergent (% of total nitrogen)
NDFCP = Neutral detergent fiber corrected for CP
Lig = Lignin
(As)= Ashes
EE= Ether extract. 

Digestible energy (DE Mcal/ kg), metabolic energy (ME Mcal/kg) and net lactation energy (NEL Mcal/ kg) for dairy cows, were estimated from TDN (NRC 2001). 

Statistical analysis 

Treatments were assigned to experimental units (silos) using a completely randomized design with four replications. The molasses and LAB treatments were applied in a factorial arrangement. Significance of treatment effects and the interaction were assessed by analysis of variance using the GLM procedure of SAS (1985). Treatment means were compared using Duncan’s multiple range test. Trends associated with molasses application rates were assessed using linear regression. 
 

Results and discussion 

Soybean forage 

Dry matter concentration of the soybean herbage harvested at the R6 stage of development averaged 24.8 % (Table 1).

Table 1.  Dry matter (DM), crude protein (CP), neutral detergent-soluble carbohydrates (NDSC) and buffer capacity (BC) of soybean forage harvested at the R6 (full seed) stage of development

 

DM

CP, %  DM

NDSC, %  DM

BC, meq./kg DM

Soybean forage (R6)

24.8 ±1.6

20.2 ±2.2

25.4±2.3

515.2±113.7

Values represent the mean of four replications ± a confidence interval  (P£0.05)

It is generally recommended (Catchpoole and Henzel 1991), that DM should be greater than 30% in the fresh forage to inhibit growth of Clostridium and prevent secondary fermentation. DM content can be increased with partial dehydration or adding molasses or mixing with forages of greater DM. However, partial dehydration should be considered carefully, because the concentration of DM may be accompanied by a sacrifice in total DM, because of a CO2 loss due to forage respiration, that may vary from 2 to 8% and even more in field wilted forage (Moser 1995).  

Average CP content of the soybean forage was 20.2% DM, which is significantly higher than that produced by tropical grasses. From a nutritional perspective, its consumption by bovines would serve to reduce the investment in imported grain proteins that are used to prepare balanced diets.  

Soybean forage had 25.4% NDSC. The NDSC are constituted by non-fiber carbohydrates (organic acids, sugars and starch) and by a soluble fiber fraction represented by fructosans, pectins and b-glucosans (Hall 1998). In spite of having this concentration of NDSC, similar to most legumes, soybean forage has a relatively low concentration of water soluble carbohydrates (sugars). In general, this is due to the fact that i) the major reserve carbohydrate in soybeans is starch, which is insoluble in water (Smith 1973); ii) pectins are abundant in soybean tissues (7 to 14%) (Chesson and Monro 1982) and iii) soybean forage has a high buffer capacity, 515 meq/kg DM), due to the high concentration of proteins and organic acids (malic, citric, nicotinic, malonic and gliceric) and their salts presents in its tissues (Playne and McDonald 1966). Mc Donald (1981) reported buffer capacity values of 488 and 540 meq/kg DM for alfalfa and red or white clover, respectively, which were similar to those found for soybean in our analysis. Apparently, CP contributes from 10 to 20% only, of this relatively high buffer capacity of legume tissues. 

Our results indicate that soybean forage is similar in composition to other legumes and is likely difficult to ferment because of its relatively low DM and soluble sugar content and its high buffer capacity. Therefore, a supplemental source of soluble carbohydrates may be needed in order to obtain a good quality silage made from soybean herbage. 

Sensory analysis of soybean silage 

Treatments containing a molasses concentration of 6% of forage fresh weight received the highest composite sensory score (76) (Table 2).

Table 2.  Sensorial grading of soybean silage with different concentrations of molasses and inoculated or not with Lactobacillus brevis 3

Parameter

Without Lb3

With Lb3

Molasses, % of fresh weight

Molasses,  % of fresh weight

0

3

6

9

0

3

6

9

Odor

18

36

54

54

18

36

54

54

Texture

22

22

22

11

22

22

22

11

Total

40

58

76

65

40

58

76

65

Note: Color was not evaluated because molasses masks natural silage color tone (Ojeda et al 1991).

Treatments with 0% molasses, scored less than 40, thus demonstrating that soybean forage can not be ensiled without an additional source of soluble carbohydrates. Improvements in silage using sugar cane molasses have been reported by others. Vallejo (1995) reported improved fermentation of silage made from tropical tree leaves with addition of molasses. León and Montenegro (2001) and Tobía et al (2003) reported improvements in soybean silage quality when mixed with corn forage and molasses, respectively. 

Inoculation with Lb3 did not affect odor nor texture (Table 2). However, sensory analysis as proposed by Ojeda et al (1991) although subjective, can be an effective tool to evaluate silage quality in the field, in a fast, simple and economic way, as was confirmed with our evaluation (Table 2). This sensory test is highly repeatable and becomes particularly valuable when it is associated with a chemical analysis of the silage. 

Soybean silage fermentation  

Silage pH decreased inversely (P£ 0.0001) to molasses concentration (Figure 1; Table 3).



Figure 1.
  Effect of molasses concentration and the inoculation
with Lactobacillus brevis 3 on the soybean silage pH



Table 3.  Response of different variables associated with the fermentation process of soybean silage to different concentrations of molasses and inoculated
or not with Lactobacillus brevis 3

Parameter 

 Without Lb3

With Lb3

 ANOVA

 % Molasses

0

3

6

9

0

3

6

9

Molasses

Lb3

Mol*Lb3

pH

5.81 e

4.50 d

4.03 a,b

4.01 a,b

5.95 f

4.41 c

4.08 b

3.98 a

P≤0.0001

ns

P≤0.03

NH3,%Total Nitrogen

21.1 f

12.6 d

7.8 b

6.4 a

19.7 e

10.8 c

8.1 b

6.0 a

P≤0.001

P≤0.013

ns

Acetic acid, %DM

2.15a

1.60b,c,d

1.40 d

1.92 a,b

1.30 d

1.80 a,b,c

1.22 d

1.84 a,b,c

P≤0.0029

ns

P≤0.02

Lactic acid, %DM

0.00 d

4.40 c

6.55 a

5.57 a,b,c

0.00 d

4.57 b,c

5.70 a,b,c

6.40 a,b

P≤0.001

ns

ns

Butyric acid, %DM

2.61 a

0.06 c

0.00 c

0.00 c

2.36 b

0.00 c

0.00 c

0.00 c

P≤0.001

P≤0.03

P≤0.04

Treatment means in a row with different letters  are statistically different (P≤0.05) according to the W. Duncan test. ns=(P≥0.05) 

Lowest pH values were obtained when this additive was used at 6 and 9%. These values are within the range recommended by Muck (1988) (3.9 to 4.2) for good quality silage. Similar results to these were also found in a previous experiment (Tobía et al 2003) when molasses was used in concentrations of 4 and 8% of soybean forage fresh weight, and by Singh et al (1996) and by Ojeda and Montejo (2001) in alfalfa and white mulberry shrub silages, respectively. Kizilsimsek et al (2005) also reported pH values of 4.2 and 4.3 in soybean silage mixed with corn and sorghum, respectively, showing that other sources of carbohydrates, perhaps more accessible than molasses can be used instead. In fact, a soybean/corn silage in a 35/65 mixture has shown promising results in dairy and dual-purpose farms in Lara and Mérida, Venezuela (Tobía et al 2007).  Those pH values found in micro silos with 0 and 3% molasses in this research where above that suggested range, regardless of inoculation with Lb3. 

Ammonia decreased (P£ 0.0001) with increasing molasses concentration for both the inoculated and control silages (Figure 2).



Figure 2.
 Effect of molasses concentration and the inoculation
with Lactobacillus brevis 3 on the soybean silage NH3/%TN content


Lowest concentrations (6.0 and 6.4 NH3/%TN) were found in silage with 9% molasses (Table 3). Optimally, NH3 values of silage should be less than 7 (Ojeda et al 1991). Our results are similar to those reported by Ojeda and Montejo (2001) in white mulberry, where NH3 decreased to 12.1, 10.5 and 7.5, when molasses was added at rates of 2.0, 4.0 and 6.0% respectively. Silages to which no molasses was added had NH3 values higher than 20% DM, which would be undesirable (Table 3). 

The usefulness of molasses in the fermentation of soybean silage was also confirmed by the increase in total organic acid concentration (Table 3). Adding molasses increased lactic acid, LA, (P£ 0.001), and to decreased butyric acid, BA, (P£ 0.001) and acetic acid, AA, (P£ 0.0029). This fermentation pattern is typical when soluble carbohydrates have been added to other silages (McDonald 1981 and Vallejo 1995). The best overall response in relative and absolute concentrations of AA, BA and LA was found in silages treated with 6% molasses (Table 3). These silages had AA and BA values lower than 1.8 and 0.1% of DM, which represent optimum values for these acids in a silage, according to Ojeda et al (1991). Those silages treated with 6 and 9% molasses reached LA contents between 5 and 10% of DM (Table 3), also considered as a good indicator of a good silage (McDonald 1981). Owens et al (1999) in a study conducted for three years on alfalfa silage previously dehydrated to 35% DM and harvested at different times during the day, reported average values of 1.7, 0.04 and 7.0% of DM, for AA, BA and LA, respectively. These values are very close to our results with soybean inoculated with Lb3 and containing 9% molasses (Table 3). Broderick et al (2002) also working with alfalfa reported AA, BA and LA contents of 2.75, 0.46 and 4.35%, respectively. 

Inoculation with Lb3 was not as important as molasses in the fermentation process as well as in the improvement of silage sensorial indicators. Lb3 did not decrease pH nor AA and did not increase LA. However, this additive decreased NH3 (P£ 0.013) as also reported by McAllister et al (1998), who found a reduction of this component in alfalfa silage inoculated with Lactobacillus plantarum (LP) + Streptococcus faecium and LP alone. Inoculation with Lb3 also reduced BA in the silage (P£ 0.03) (Table 3). 

A molasses x Lb3 interaction was found in the reduction of AA (P£ 0.02). However, this additive effect is not clear since a concentration of 3% molasses caused an increase of 11% in AA, but this acid decreased 13 and 4% in silages treated with 6 and 9% molasses, respectively (Table 3). 

Silage analysis 

Adding molasses to the soybean herbage increased (P£ 0.001) DM concentration of the ensiled material (Table 4).

Table 4.  Chemical composition of soybean silage in response to different concentrations of molasses and inoculated or not with Lactobacillus brevis 3

Parameter

Without Lb3

With Lb3

ANOVA

Molasses,  %

0

3

6

9

0

3

6

9

Molasses

Lb3

Mol*Lb3

Dry matter, %

25.1 f

27.1 e

29.8 c

31.0 b

24.5 g

28.3 d

29.9 c

31.6 a

P≤0.001

P≤0.014

P≤0.001

Crude protein, % DM

21.7 a

19.4 b,c

19.0 b,c

18.2 c

21.4 a

18.9 b,c

19.8 b

18.6 b,c

P≤0.0001

ns

ns

Ether extract, % DM

11.1 a

10.3 a

8.0 b,c

7.0 d

10.6 a

8.6 b

7.6 c,d

6.0 e

P≤0.0001

P≤0.0018

ns

Ash, % DM

8.5 c

8.7 c

9.6 b

10.0 a,b

8.4 c

8.6 c

9.7 a,b

10.1 a

P≤0.0001

ns

ns

Treatment means in a row with different letters  are statistically  different (P≤0,05)  according to the  W.Duncan test.  ns=P≥0,05)

This was due to the higher DM concentration of molasses (75%), as compared to that of the soybean forage (25%). This observation agrees with results of Ojeda and Montejo (2001) in white mulberry silage and Vallejo (1995) in tropical woody tree leafage silage. In the latter study there was 6.5% increase in DM with the addition of 5% molasses. Similar results have been reported when other sources of soluble carbohydrates like heart of palm fruits (Rojas et al 1998) or cassava roots (Sujatha et al 1986) have been added to the silage. This increase in DM is beneficial because it reduces nutrient effluxes from the silos and favors silage palatability. With no exception found so far, the use of molasses ameliorates silage sensory characteristics and consumption. Congruent with the DM increase was the increase (P£ 0.001) observed in the ash concentration in the soybean silage (Table 4). The ash concentration in this additive (13.3%, according to NRC (2001) is 1.6 times that of soybean forage (ca. 8.5%). Ash concentration varies with soil fertility, fertilizer application, soil pH, and other soil and climatic variables, but soybean forage ash concentration observed in soybean silage fermented with 9% molasses (Table 4) is quite similar to that found in alfalfa (Broderick et al 2002).  

Molasses reduced (P£ 0.0001) CP and EE content in soybean silage (Table 4), which is interpreted as a dilution effect of this fat and protein free additive. However, CP content of soybean silage in our experiment, even in silages treated with 6 and 9% molasses, was relatively high as compared to those found by Pogue and Arnold (1979). On the other hand, most reports for CP in alfalfa silages (Weiss and Shokey 1991, Dhiman and Setter 1997, McAllister et al 1998 and Broderick et al 2002) are very similar to ours (ca. 20%) confirming again the similarity in protein value between these two legumes . 

The reduction in EE induced by the addition of molasses to the soybean silage is of little relevance, in a silage that contains almost 200% more EE than that of alfalfa (aprox. 7.5% of total DM, in silages containing 6 and 9% molasses). The relatively high EE content of soybean silage is an important property, since this represents a high energy contribution to balanced formulas for animal feed. However, it should be noted that the final EE content in concentrate rations for dairy cows should not exceed 5% (Palmquist and Jenkins 1980). Wiederholt y Albretcht (2003) recommend that soybean silage should not exceed 50% in terms of dry matter content in a ration, to avoid its rejection by bovines.  

Lb3 inoculation increased (P£ 0.014) DM in soybean silage. In this aspect, our results resemble those found by Singh et al (1996) in alfalfa silage inoculated with Lactobacillus plantarum or Estreptococcus faecalis and with the combination of both microorganisms. This response can be attributed to the suppressive effect of non beneficial microorganisms (i.e. Clostridium) by LAB. Some microorganisms degrade CP and other DM components to produce volatile compounds like NH3 and CO2. On the other hand, there was no effect of Lb3 on the CP and ash content of the silage, whereas it caused a small reduction (P£ 0.001) in EE.    

An Lb3 x molasses interaction (P£ 0.001) for DM occurred. The greatest increase (4.4%) in DM was for silage treated with molasses at 3% (Table 4)

Cell wall structural components and NDSC 

Concentrations of cell wall components (NDF, ADF and cellulose) decreased linearly with increasing concentrations of molasses. Again, a dilution effect due to the addition of this non fiber additive (0.4% NDF vs. 40% in the soybean forage, for instance) partially explains this response. However, the relatively abrupt drop in those components, at least in NDF where a molasses x Lb3 interaction was found (P£ 0.04) (Table 5), may also involve cell wall degradation, since NDF reduction in silages treated with molasses was 26% whereas without molasses this reduction was only 21%.  

McAllister et al (1998) also found a reduction of 5 percentage points in NDF silage inoculated with Lactobacillus plantarum. A decrease in lignin concentration, although supported by the statistical analysis (P£ 0.003), was of small magnitude.

Table 5.  Cell wall constituents and neutral detergent-soluble carbohydrates of soybean silage in response to different concentrations of molasses and inoculated
or not with Lactobacillus brevis 2

Parameter, % dry matter

Without Lab3

With Lab3

ANOVA

Molasses, % 

0

3

6

9

0

3

6

9

Molasses

Lb3

Mol*Lb3

Neutral detergent fiber

42.6 b

36.8 c

32.5 d

31.9 d

44.7 a

35.8 c

32.5 d

31.2 d

P≤0.0001

ns

P≤0.04

Acid detergent fiber

37.0 a

30.9 b

26.7 c

27.3 c

37.8 a

30.6 b

26.7 c

26.8 c

P≤0.0001

ns

ns

Cellulose

30.8 a

24.8 b

21.3 c

21.0 c

30.8 a

24.9 b

21.0 c

20.5 c

P≤0.0001

ns

ns

Hemicellulose

5.6 a

5.9 a

5.8 a

4.6 a

6.9 a

5.2 a

5.8 a

4.4 a

ns

ns

ns

Lignine

7.4 a

6.5 a,b

6.2 b,c

6.1 b,c

7.3 a

5.7 b,c

6.4 a,b,c

5.8 b,c

P≤0.003

ns

ns

NDSC, %

18.4 e

27.1 d

33.4 b

35.3 a,b

17.2 e

30.6 c

33.3 b

36.7 a

P≤0.0001

ns

ns

Treatment means in a row with different letters   are statistically different (P≤0.05) according to the W.Duncan test. ns=P0.05
DM=Dry matter; NDF=; ADF=; NDSC=Neutral detergent soluble carbohydrates

It is worth mentioning that NDF values of soybean silage without molasses are similar to those values found in alfalfa (Weiss et al 1991 and Broderick et al 2002), and lower than those NDF values found in grasses which are considered forages of lower nutritional value than legumes (Weiss et al 1991)

The NDSC content in the soybean silage increased linearly (P£ 0.0001) with the addition of molasses (Figure 3; Table 5).



Figure 3. 
Effect of molasses concentration on the cell wall constituents and
the neutral detergent-soluble carbohydrates content in the soybean silage


However, molasses not only contributed as a source of soluble carbohydrates but also as a way to prevent NDSC loss from the silage, since those silages made without molasses showed values of 17 and 18% NDSC (Table 5), whereas that of the soybean forage (Table 1) was 25.4 %. Similar results have been observed in alfalfa silage (McDonald 1981). Part of this loss might take place during that period that goes from forage harvesting till the onset of the fermentation process.  

However, NDSC values (6.1%) much lower than those found in this research were reported in alfalfa silage under a relatively deficient fermentation (pH 5.05 and NH3/ %TN 19.1) (McDonald 1981). 

When those values shown in Table 1 are compared with those in Tables 4 and 5, a general statement on the benefits of silage as an inexpensive and simple way to efficiently preserve the nutritional properties of the soybean forage is strongly supported.  

Energy concentration in the soybean silage  

The best energy concentration, as indicated by the value of total digestible nutrients, (TDN1x), was found in those silages containing 3% molasses, regardless of Lb3 inoculation (Table 6).

Table 6.  Energy concentration in soybean silages in response to molasses concentrations and inoculation with Lactobacillus brevis 3

Units

Without Lb3

With Lb3

Molasses, %

0

3

6

9

0

3

6

9

TDN1x, % DM

69.41 c,d

72.15 a

70.20 b,c

68.86 c,d

68.01 d

71.64 a,b

69.45 c,d

67.99 e

TDN3X, % DM

65.03

66.77

65.52

64.67

64.13

66.45

65.05

64.11

DE1X, Mcal.kg DM-1

3.06

3.18

3.10

3.04

3.00

3.16

3.06

3.00

DE3X, Mcal.kg DM-1

2.87

2.94

2.89

2.85

2.83

2.93

2.87

2.83

ME3X, Mcal.kg DM-1

2.45

2.52

2.47

2.43

2.41

2.51

2.45

2.41

NEL3X, Mcal.Kg DM-1

1.47

1.52

1.49

1.46

1.45

1.51

1.48

1.45

Treatment means in a row with different letters are statistically different (P≤0,05), according to the W.Duncan test. TDN1x% (Weiss 1999; NRC 2001)

TDN3X  %=TDN1X %  X  discount =[(TDN1X % - ((0,18  X TDN1X %) – 10.3) X number of times of DM con sumption))] / TDN1X %, NRC 2001

DE1X (Mcal . kg DM-1) = 0.04409 . TDN1X%   (NRC 2001)

DE3X (Mcal . kg DM-1) = 0.04409 . TDN3X%  

ME3X = 1.01 . DE3X (Mcal . kg DM-1) – 0.45  (NRC 2001)

NEL3X (Mcal . kg DM-1) =0.0245 . TDN3X – 0.12   (NRC 2001)

1X = DM consumption (DMC), non-lactating cows (maintenance)

3X = DMC (double maintenance cows), lactation  cows

TDN = total digestible nutrients; DE = digestible energy; ME = maintenance energy

NEL = net lactation energy
Note : DE, ME and NEL  were estimated from TDN determinations.                     

The relatively high energy concentration values in the soybean silage are due in part to the high EE, which is a component (EFA) used to estimate TDN; other components of soybean DM are similar to those found in alfalfa silage and should not make a difference in TDN.  

Net lactation energy (NEL) values of soybean silage were 12.5 % higher than those reported by Weiss et al (1991) in alfalfa and 25% higher than those reported by the NRC (2001) for immature alfalfa (40 to 46% NDF). In corn, ENL values similar to those found in our research (Table 6), have been reported (NRC 2001). 
 

Conclusions 

Acknowledgements 

We want to express our acknowledgements to the Government of Costa Rica for the financial support to conduct this research through the “Programa de Fondos Concursables”. We want to thank Ing. Alberto Quintana, his son Alberto and the field and dairy workers at the Hacienda Pozo Azul, for their cooperation, responsibility and enthusiasm during the field work involved in this research. We also thank Jesús Calvo, Administrative Head of the Grain and Seed Research Center (CIGRAS) of the University of Costa Rica, the Consejo Nacional de Investigaciones Científicas y Tecnológicas of Costa Rica (CONICIT) and the Fundación de la Universidad de Costa Rica para la Investigación (FUNDEVI) for their efficient fund accounting and controlling work. We also appreciate the Directors of CIGRAS, the Animal Nutrition Research Center and the Agronomy Research Center of the University of Costa Rica, for allowing us to use their laboratories and other facilities to conduct different aspects of this project. 
 

Literature cited 

AOAC 1990 Association of Official Analytical Chemists. Official Methods of Analysis, 15th Edition, Washington D.C., 1298 p.

Blount A R S, Wright D L, Sprenkel R K, Hewitt T D and Myer P O 2006 Forage soybeans for grazing, hay and silage. University of Florida, IFAS Extensión. Retrieved January 4, 2008 from http://edis.ifas.ufl.edu/AG184

Bolsen K, Brent, B and Uriarte E 2001 The silage triangle and important practices often overlooked. California Animal Nutrition Conference. California, USA, pp. 60-65

Broderick G, Koegel R, Walgenbach R and Kraus T 2002 Ryegrass or alfalfa silage as the dietary forage for lactating dairy cows. Journal of Dairy Science 85:1894-1901 http://jds.fass.org/cgi/reprint/85/7/1894

Catchpoole V and Henzel E 1991 Silage and silage making from tropical herbage species. Herbage 41: 213-221

Chesson A and Monro J 1982 Legume pectin substances and their degradation in the ovine rumen. Journal of Science of Food and Agriculture 33: 852

Dhiman T and Satter L 1977 Yield response of dairy cows fed with different proportions of alfalfa silage and corn silage. Journal of Dairy Science 80: 2069-2082 http://jds.fass.org/cgi/reprint/80/9/2069

Fehr W and Caviness C 1980 Stages of soybean development. Iowa Cooperation Service, Special Report 80

Goering H and Van Soest P 1970 Forage fiber analyses. Agriculture Research Service USA, Handbook 379

Hall M 1997 Interpreting Feed Analysis: Uses, abuses and artifacts. Eight Annual Florida Ruminant Nutrition Symposium, Gainsville, Florida. Retrieved on November 12, 2007 from http://www.animal.ufl.edu/hall

Hall M 1998 Making nutritional sense of non-structural carbohydrates. Ninth Annual Florida Ruminant Nutrition Symposium, Gainesville, Florida. Retrieved on November 12, 2007 from http://www.animal.ufl.edu/hall

Hintz R. and Albretcht K 1994 Dry matter partitioning and forage nutritive value of soybean plant components. Agronomy Journal 86: 59-62

Jaster E 1995 Legume and grass silage preservation. In: Moore K J, Peterson M, Kral D and Viney M (Editors), Post-harvest Physiology and Preservation of Forages. ASA, CSSA, Madison, Wisconsin, pp 91-115

Kizilsimsek M, Erol A and Calislar S 2005: Effects of row material and silo size on silage quality. Livestock Research for Rural Development  Volume 17 Art #33 Retrieved February 11, 2008 from  http://www.lrrd.org/lrrd17/3/kizi17033.htm

León B and Montenegro A 2001 Sistema de cultivo de maíz (Zea mays) asociado con soya (Glycine max) para la elaboracion de ensilaje. “Licenciatura” Thesis, School of Animal Science, University of Costa Rica, 90 p

McAllister T R, Feniuk Z, Mir P, Mir L, Selinger L and Cheng K 1998 Inoculant for alfalfa silage: Effect on aerobic stability, digestibility and the growth performance of feedlot steers. Livestock Production Science 53: 171-181

McDonald P 1981 The biochemistry of silage. First Edition, John Wiley and Sons, New York, USA, 226 p

Moser L E 1995 Post-harvest physiological changes in forage plants. In: Moore K J, Peterson M, Kral D and Viney M (Editors), Post-harvest Physiology and Preservation of Forages. ASA, CSSA, Madison, Wisconsin, pp.1-19

Muck R 1988 Factors influencing silage quality and their implication for management. Journal of Animal Science 71: 2992-3002 http://jds.fass.org/cgi/reprint/71/11/2992.pdf

Muck R and Kung L 1997 Effects of silage additives on ensiling proceedings from the silage: field to feedbunk. North American Conference. Hershey, Pennsylvania, USA, pp 187-189

NRC (National Research Council) 2001 Nutrient requirement of dairy cattle. Seventh revision. Edition, National Academy Press. Washington D.C., 381 p

Ojeda F and Montejo I 2001 Conservación de la morera (Morus alba) como ensilaje. I. Efecto sobre los compuestos nitrogenados. Pastos y forrajes 24: 147-55

Ojeda F, Cáceres O and Esperance M 1991 Conservación de forrajes. Edición Pueblo y Educación, Cuba, 80 p.

Owens V, Albrecht K, Muck R and Duke S 1999 Protein degradation and fermentation characteristic of red clover and alfalfa silage harvested with varying levels of total nonstructural carbohydrates. Crop Science 39: 1873-1880

Palmquist E and Jenkins T 1980 Fat in lactation rations: Review. Journal of Dairy Science 63: 1-14 http://jds.fass.org/cgi/reprint/63/1/1

Playne M and McDonald P 1966 The buffering constituents of herbage and of silage. Journal of Science of Food and Agriculture 17: 264-268

Pogue E and Arnold B 1979 Corn-soybean silage compared with corn silage for milk production. Research Report. Misissippi Agricultural and Forestry Experiment Station 17: 1-3

Rojas A 1985 Effect of rolled corn silage on digestion of nutrients and feedlot performance of growing steers. MSc.Thesis Department of Animal Science, Iowa State University Library, 93 p

Rojas A, Ugalde H and Aguirre D 1998 Efecto de la adición de fruto de pejibaye (Bactris gasipaes) sobre las características nutricionales del ensilaje de pasto gigante (Pennisetum purpureum). Agronomía Costarricense 22: 145-151 http://www.pejibaye.ucr.ac.cr/NutricionAnimal/Nutricion%20animal2.htm

SAS (Statistical Analysis System) 1985 Guide for personal computer. Fifth Edition, SAS Institute. North Carolina, USA

Singh A, Edward J, Mor S and Singh K 1996 Chemical and microbiological change during ensiling of lucerne with lactic acid bacteria and molasses. Indian Journal of Dairy Science 49: 593-598

Smith D 1973 Chemistry and biochemistry of herbage. In: Butler G and Bailey R (Editors), Academic Press, New York, USA, Volume I pp 106-155

Sujatha A, Fontenot J and Jarasuviya M 1986 Ensiling characteristics of tropical grasses as influenced by state of growth, additives and chopping length. Journal of Animal Science 63: 197-207 http://jas.fass.org/cgi/reprint/63/1/197

Tobía C, Sequera C, Cioffi R, Villalobos E and Escobar O 2007 Experiencias en la elaboración de silaje maíz-soya en dos sistemas de producción bovina en Venezuela. pp.78-87 In: Trejos R, Zambrano C, García W, Tobía C, Mancilla L, Valbuena N J and Ramírez F (Editores). Seminario sobre el manejo y utilización de pastos y forrajes en sistemas de producción animal. Barquisimeto, Venezuela. 145p.

Tobía C, Uribe L, Villalobos E., Soto H and Ferris I 2003 Aislamiento, selección y caracterización de bacterias ácido lácticas en ensilajes de soya (Glycine max L. Merr.). Agronomía Costarricense 27: 21-27 http://www.mag.go.cr/rev_agr/v27n02_021.pdf

Tobía C and Villalobos E 2004 Producción y valor nutricional del forraje de soya (Glycine max L. Merr.) en condiciones tropicales adversas. Agronomía Costarricense 28: 17-25 http://www.mag.go.cr/rev_agr/v28n01_017.pdf

Vallejo M 1995 Efecto del premarchitamiento y la adición de melaza sobre la calidad del ensilaje de diferentes follajes de árboles y arbustos tropicales. MSc. Thesis, Centro Agronómico Tropical de Investigación y Enseñanza (CATIE), Turrialba, Costa Rica. 117 p

Villalobos E and Camacho F 2003 Registration of ‘CIGRAS-06’ soybean. Crop Science 43: 1122

Weiss W 1999 Energy prediction equation for ruminant feeds. In: Proceedings Cornell Nutrition Conference, New York, USA, pp 176-185

Weiss W and Shockey W 1991 Value of orchard grass and alfalfa silage fed with varying amounts of concentrates for dairy cows. Journal of Dairy Science 74: 1933-1943 http://jds.fass.org/cgi/reprint/74/6/1933.pdf

Wiederholt R and Albretcht K 2003 Using soybean as forage. Focus on forage. University of Wisconsin Extension. WI. Retrieved February 11, 2008 from:  http://www.uwex.edu/ces/crops/uwforage/SoybeanForage.htm



Received 7 January 2008; Accepted 12 February 2008; Published 3 July 2008

Go to top