The effects of Saccharomyces cerevisiae strains on the rumen fermentation in sheep fed with diets of different forage to concentrate ratios

Abstract

The objective of this study was to compare the effects of two different Saccharomyces cerevisiae yeast strains on the rumen fermentation of sheep fed with two different total mixed rations. Three rumen-cannulated Merino wethers were used in a pilot study in which a phase of feeding a higher proportion of concentrate in the diet (HC; forage to concentrate ratio 40:60, as fed) was followed by a phase of lower proportion of concentrate in the diet (LC; forage to concentrate ratio 49:51, as fed). Both phases consisted of three 3-week periods as follows: no yeast supplementation, trehalose non-producing and trehalose-producing S. cerevisiae supplementations, respectively. Rumen fluid samples taken in the last days of each period were measured for pH, short chain fatty acid (SCFA) concentrations and ammonia content. The degradability of neutral detergent fibre (NDF) and starch was estimated by the in sacco method. Supplementation with the trehalose non-producing strain did not alter any variables tested in the HC phase. In the LC phase, though degradability was not altered, total SCFA concentration increased, resulting in a decrease of rumen fluid pH. The trehalose-producing strain increased total SCFA content and effectively prevented the consequent decrease in pH in both phases. These effects were more pronounced in the high-concentrate phase, as shown by higher molar proportion of propionate, increased level of ammonia content, and higher ruminal degradability of NDF and starch compared to control and the trehalose non-producing yeast strain.

Keywords:

• Saccharomyces cerevisiae
• trehalose
• rumen fermentation

View correction statement:

Erratum

This article was originally published with errors. This version has been corrected. Please see Erratum (http://dx.doi.org/10.1080/09712119.2013.889900)

1. Introduction

Viable Saccharomyces cerevisiae yeast strains are widely used as additives in ruminant nutrition for their favourable impact on animal production. However, the responses to yeast culture supplementation are highly variable and apparently infiuenced by the composition of the diet (Chaucheyras-Durand et al. 2008; Desnoyers et al. 2009; Opsi et al. 2011), or which yeast strain is applied (Newbold et al. 1995; Kutasi et al. 2004; Könyves et al. 2007; Chaucheyras-Durand et al. 2008). It has been shown that different yeast strains adapt differently to rumen conditions. The S. cerevisiae does not grow in rumen fluid (Arambel & Tung 1987), but retains its metabolic activity (Ingledew & Jones 1982) and viability (El Hassan et al. 1994). The presence of trehalose and sucrose can alter the protein structure of the cell membrane in certain bacteria and thus increase the viability of cells in the dried state (Leslie et al. 1995; García de Castro et al. 2000). Our previous in vitro study (Kutasi et al. 2005) has come to the conclusion that trehalose accumulation in yeast cultures increases their viability in the dried state, as proven by a higher number of colony forming units (CFU) when rehydrated, compared to trehalose non-producing control. The aim of the present study was to determine whether S. cerevisiae yeast strains that produce trehalose in a higher amount – and thus supposedly show higher metabolic activity when rehydrated in the rumen – affect rumen fermentation differently than trehalose non-producing strains. Another objective was to study the influence of the forage to concentrate ratio of diets on the effects of yeast supplementation. Evaluating the effects of different yeast supplementation in both diets could provide information on which preparation could be used effectively in ruminant nutrition.

2. Materials and methods

2.1. Animals and sampling

Three rumen-cannulated Merino wethers (body weight = 70 kg) were used in this pilot study. Two experimental total mixed rations (TMRs), differing in their forage to concentrate ratio, were formulated. The experiment consisted of a phase of feeding a diet with a higher proportion of concentrate (HC), followed by a phase of a lower proportion of concentrate in the diet (LC). The TMRs (Table 1) consisted of corn silage, meadow hay, barley grain, corn grain, extracted sunflower meal and extracted soybean meal, meaning 60% concentrate and 40% forages in the HC diet (as fed) and 51% concentrate and 49% forages in LC diet (as fed). The daily amount of TMR was 1.2 kg/d/animal (on a dry matter basis). Both phases (HC and LC) consisted of three 3-week periods during which animals received: no yeast supplementation (control, CO), 2.5 g/d dried preparation of trehalose non-producing S. cerevisiae NCAIM 001262 (Treatment 1, T1), and 2.5 g/d dried preparation of trehalose-producing S. cerevisiae NCAIM 001263 (Treatment 2, T2), respectively. The yeast preparation was mixed in the daily amount of TMR which was fed in two equal portions at 8.00 am and 4.00 pm.

Table 1. Ingredients and chemical composition of the diets.

Composition	HC diet	LC diet
Ingredient (as fed)
Corn silage (%)	54	47.5
Meadow hay (%)	6.0	16.0
Barley (%)	11.5	5.5
Corn (%)	15.0	16.0
Extr. sunflower meal (%)	5.8	7.7
Extr. soybean meal (%)	6.2	5.8
Limestone (%)	1.0	1.0
Premix (%)^a	0.5	0.5
Chemical composition
Dry matter (g/kg)	588	622
	in 1 kg dry matter
NEm (MJ)^b	7.2	6.8
Crude protein (g)	162.4	162.3
Metabolizable protein energy (MPE, g)^b	109.3	106.6
Metabolizable protein nitrogen (MPN, g)^b	106.2	106.1
Crude fibre (g)	130.7	166.5
Neutral detergent fibre (NDF, g)	303.9	350.1
Acid detergent fibre (NDF, g)	147.2	184.5
Ether extract (g)	28.3	28.9
Ca (g)	4.0	4.1
P (g)	3.1	3.2

HC diet, higher proportion of concentrate; LC diet, lower proportion of concentrate.

^a Contained: Calcium, Manganese, Zinc, Iron, Selenium, Iodine, Cobalt, Vitamins (A, D3, E, B1, B2, niacin, folic acid, pantothenic acid, B6, B12, cholin-chloride), anti-oxidants.

^b NE_m, MPE and MPN are calculated values.

Between T1 and T2 periods, a one-week rest, between HC and LC phases, two weeks of rest were held, during which animals received only meadow hay, to avoid any influence of the prior treatment or diet. The first two weeks of experimental periods served as an adaptation for the rumen environment. The samples were taken in the third week. As controlled by regular analysis, the composition of feeds did not deteriorate in any way during the course of the experiment. The animals were kept in a climate-controlled room. The daily ration met the maintenance requirement, and was totally consumed by every animal on every day of the experiment. By applying such measures in the feeding of the animals, there were no differences in dry matter intake that could alter rumen fermentation.

The rumen fluid samples were taken on days 15, 17 and 19 of each period, three hours after the morning feeding. The rumen fluid pH was measured immediately after sampling by a digital pH meter (Radelkis OP-211/1, Radelkis Co., Budapest, Hungary). The ruminal fluid samples were squeezed through four layers of gauze and then centrifuged (1000 g, 20 min, at room temperature). One mL of 25% HPO₃ was added to 5 mL of filtered rumen fluid for short chain fatty acid (SCFA) determination (Supelco Inc. 1975). The acidified fluid was centrifuged (1000 g, 20 min, at room temperature) and the supernatant was stored at −20°C until analysis. The ruminal SCFAs were separated and quantified by gas chromatography (Shimadzu GC 2010, Japan) using a 30 m (0.25 mm i.d.) fused silica column (Nukol, Cat.# 24107, Supelco, Bellefonte, PA). The ammonia content of ruminal samples was determined by the method of Chaney and Marbach (1962). The degradability of neutral detergent fibre (NDF; alfalfa pellet) and starch (corn and barley) was estimated by using the in sacco method in the last week of each period. Seven incubation periods (2, 4, 8, 12, 24, 48 and 72 h, respectively) were used, with two bags for each time interval for each sheep. Degradability constants a, b and c for NDF were derived from the following equation: proportional NDF disappearance = a + b(1 − e−ct), where a = the soluble fraction, b = the potentially degradable fraction not including a, and c = the fractional rate of degradation per hour of the b fraction with time t (Ørskov & McDonald 1979). The effective degradability (p) at a given fractional ruminal outflow rate per hour (r) was calculated from the equation p = a + (b × c)/(c + r). The NDF content was measured according to the procedure described by Robertson and Van Soest (1981). In the case of starch degradation, test samples were incubated for 8 h in the rumen. Starch content of barley and corn grain samples was measured by a polarimetric method according to the Commission Directive 1999/79/EC (The Commission of the European Communities 1999).

2.2. Statistical analyses

Considering the fact that the same animals were the subject of multiple measurements a linear mixed model was applied for statistical analyses. The fixed effects were the treatment, the diet and the sampling; the random effects were the sheep and the day.

Model effects were tested together based on their F values. All factors and potential interactions were evaluated with the cut-off for inclusion being p < 0.05. Estimation of treatment differences was calculated using contrast matrices. The p-values were adjusted by Tukey–Kramer correction at multiple testings. The statistical analysis was performed by the statistical software R (R Development Core Team 2011).

3. Results

Results of the experiment are summarised in Tables 2 and 3.

Table 2. Rumen fermentation variables during the experiment (mean ± SD).

	Treatments
	HC diet			LC diet
Items	CO	T1	T2	CO	T1	T2
pH	^a15.5 ± 0.3	^a5.5 ± 0.3	^b5.7 ± 0.2	^a25.8 ± 0.2	^b5.5 ± 0.3	^a5.6 ± 0.2
Total SCFA (mmol/L)	^a91.7 ± 11.3	^a194.3 ± 14.1	^b101.2 ± 19.9	^a87.4 ± 17.8	^b2112.6 ± 17.8	^b109.7 ± 22.1
Acetate (molar %)	59.3 ± 8.7	59.4 ± 3.9	60.2 ± 2.3	62.7 ± 3.9	61.0 ± 4.8	61.9 ± 4.7
Propionate (molar %)	^a18.9 ± 5.9	^a20.5 ± 2.3	^b22.2 ± 2.1	19.1 ± 4.1	18.7 ± 2.3	20.2 ± 2.8
Butyrate, (molar %)	14.9 ± 4.1	15.8 ± 1.8	16.1 ± 1.7	15.2 ± 2.1	15.8 ± 1.8	15.4 ± 2.6
Ammonia (mg/100 mL)	^a19.7 ± 9.0	^a,b23.6 ± 8.1	^b25.1 ± 5.8	^a20.0 ± 4.7	^a22.9 ± 6.5	^b25.5 ± 3.9

Note: Means with different superscript letters within a diet are significantly different (p < 0.05). Means with different superscript numbers mark significant difference between the items in HC and LC diets (p < 0.05).

HC diet, higher proportion of concentrate; LC diet, lower proportion of concentrate; CO, control; T1, trehalose non-producing Saccharomyces cerevisiae supplementation; T2, trehalose-producing Saccharomyces cerevisiae supplementation; SCFA, short chain fatty acid.

Table 3. In sacco degradability of alfalfa pellet and cereals (mean ± SD).

	Treatments
	HC diet			LC diet
Items (%)	CO	T1	T2	CO	T1	T2
Alfalfa pellet NDF	^a119.7 ± 1.3	^a17.8 ± 2.8	^b21.7 ± 0.6	^a223.7 ± 1.1	^b20.5 ± 2.9	^b21.4 ± 1.4
Corn grain starch	53.5 ± 15.0	49.5 ± 8.6	50.7 ± 9.0	54.1 ± 7.9	50.6 ± 8.5	50.7 ± 9
Barley grain starch	^a84.9 ± 8.7	^a84.2 ± 3.9	^b193.7 ± 2.8	^a89.1 ± 2.7	^b85.9 ± 5.0	^a289.4 ± 3.4

Note: Means with different superscript letters within a diet are significantly different (p < 0.05). Means with different superscript numbers mark significant difference between the items HC and LC diets (p < 0.05).

HC diet, higher proportion of concentrate; LC diet, lower proportion of concentrate; CO, control; T1, trehalose non-producing Saccharomyces cerevisiae supplementation; T2, trehalose-producing Saccharomyces cerevisiae supplementation; NDF, neutral detergent fibre.

In the HC diet, no statistically significant difference could be observed between control and T1 treatments regarding any of the measured parameters. When feeding trehalose-producing yeast (T2), the rumen fluid pH, the total SCFA concentration and the molar proportion of propionate were higher compared to both CO and T1 values (p < 0.05). In the LC diet, the rumen fluid pH was significantly lower (p < 0.05) in the sheep fed with the diet supplemented with T1 compared to the values of T2 and control animals, respectively. Differences were observed also in the total SCFA concentrations. Values measured in the T1 and T2 periods were both higher than control values (p < 0.05), but did not differ significantly from each other. Concerning the individual SCFA concentrations in the LC diet, no significant differences were found among the treatments.

Ammonia concentration of the rumen fluid in the sheep fed with the diet supplemented with T2 was significantly higher (p < 0.05) than in the control animals where either LC or HC feeding was applied.

Comparing the effects of the two diets, we found that the pH of the rumen fluid in the control period was lower (p < 0.05) in the HC diet than in the LC diet. The total SCFA concentration in the rumen fluid was lower (p < 0.05) in the T1 supplementation period of the HC diet than in the same period of the LC diet. Other parameters showed no difference between diets (p > 0.05).

The in sacco degradability measurements revealed that by providing trehalose-producing yeast supplementation (T2) in the HC diet, the degradability of alfalfa NDF and barley grain starch improved (p < 0.05) compared to both control and T1 periods. In the periods of yeast supplementation in the LC phase, NDF degradability values were lower (p < 0.05) compared to control. The NDF degradability was significantly different between CO periods in the HC or LC diets, respectively.

The ruminal degradability of NDF was higher in sheep fed with the LC diet without any kind of yeast supplementation. The degradation rate of barley grain starch decreased in the T2 period of LC phase compared to that of HC.

4. Discussion

Yeasts as feed supplements have been the subjects of numerous studies recently (Desnoyers et al. 2009; Dehghan-Banadaky et al. 2012; Poppy et al. 2012), yet a comparison on the effects of preparations with highly different trehalose content has not been made yet. In the phase of feeding a high-concentrate diet in the present experiment, the yeast strain with no prominent trehalose-producing ability (T1) had no detectable effect on the studied rumen fermentation variables. Sauvant et al. (2004) have reported similar observations in their review on 78 experiments. Yet, in the phase of feeding a low-concentrate diet (LC), the trehalose non-producing yeast supplementation considerably increased the total SCFA concentration, resulting in a decrease in rumen pH and a consequent decline in the fibrolytic activity of rumen microbiota. Such increase in the SCFA concentrations is confirmed by others (Patra et al. 1996; Robinson 2002; Marden et al. 2008). However, Robinson (2002) and Marden et al. (2008) found in their studies that the rumen pH and the organic matter degradability increased, which explains the increase in the total SCFA concentration, as a result of the enhanced activity of pH-sensitive rumen microbes. In the present study, T1 did not considerably influence the rumen pH, and it did not change the starch or fibre degradability positively in any of the phases. The increase in SCFA concentration thus cannot be explained on the basis of available results.

When supplementing the high-concentrate diet (HC) with the trehalose-producing strain (T2) the obtained results are in accordance with the publications of the previously mentioned two authors (Robinson 2002; Marden et al. 2008), and Guedes et al. (2008) who also found that S. cerevisiae supplementation increased the SCFA concentrations and the fibrolytic activity.

As Chaucheyras-Durand et al. (2008) describe in their review, yeasts are capable of preventing the marked post-prandial decrease of rumen pH, which is generally observed after the consumption of a diet with high fermentable carbohydrate content (Nocek 1997). The sharp decline in pH is caused by an increase in the SCFA production. Similarly to our observations, Chaucheyras-Durand and Fonty (2002) found that active dry yeast supplementation maintained rumen pH at values compatible with an efficient rumen function, as shown by higher fibrolytic activity in the rumen of the supplemented animals. The buffering effect of T2 seemingly increased the activity of propionate-producing amylolytic bacteria also, as shown by higher starch degradability and higher molar proportion of propionate. Such increase in the propionate concentration is confirmed by the studies of Marden et al. (2008) and Guedes et al. (2008).

The increase in the SCFA concentration and starch degradability is in agreement with the finding of the meta-analysis of Desnoyers et al. (2009) observing that the influence of yeast supplementation is greatest when using high concentrate or high-NDF diets rather than intermediate ones. Yet, the study states that an increase in the proportion of NDF limits the pH buffering effect of yeasts, as also confirmed by our results. The in sacco degradability of NDF was lower in the HC diet compared with the LC diet. The decrease in fibre degradation can be related to the reduction in the activity of the cellulolytic microbes associated with a greater decrease in ruminal pH (Archiméde et al. 1997).

Concerning the nitrogen metabolism in the rumen, there are two variables that could be influenced by yeast supplementation and affect the ammonia concentration in the rumen. The first one is the protein degradation in the rumen, whether increased or decreased, the second one is the ammonia incorporation into microbial proteins. There is inconsistency in the relevant literature concerning the effects of yeast preparations on the ruminal nitrogen metabolism. In certain studies on the effects of yeast supplementation, a decreased ammonia concentration was observed in the rumen fluid (Erasmus et al. 1992; Moallem et al. 2009), but in others there was no difference (Chademana & Offer 1990; Corona et al. 1999; Guedes et al. 2008; Křížová et al. 2011) or even an increase in the ruminal ammonia concentration (Newbold et al. 1998). In our study, applying T2 increased the ruminal ammonia concentration, supposedly due to the increased microbial protein degradation. The increase in the ammonia concentration resulted in a higher rumen fluid pH value in case of HC diet supplemented with T2. However, the higher ammonia content in the rumen fluid of sheep fed with the LC diet with T2 supplementation was not accompanied by an increased ruminal pH. As pH is in relation with the production, absorption and flow rates of ammonia, SCFAs and other metabolites, the reason is difficult to assess. A possible cause is that the magnitude of increase in the total SCFA concentration between the control and T2 periods was lower in the HC (around 10 mmol/L) than in the LC diet (around 20 mmol/L).

When comparing the performance of the two types of strains, T2 had a significantly better effect on the rumen fermentation. The higher amount of intra- and pericellular trehalose might serve as a source of rapidly available carbohydrates, yet this is only an interesting speculation that needs further studies. More likely to be the reason behind the difference is that the induction of trehalose synthesis during the microcapsulation of T2 increases the viability of cells and the stability of the preparation in the dried state as proven by the in vitro tests (Kutasi et al. 2005). The results of the present study suggest that a more pronounced and diverse metabolic activity results in an increased number of viable cells in the ruminal microbial ecosystem, compared to the trehalose non-producing strain.

In conclusion, the trehalose-producing S. cerevisiae strain was more effective in providing favourable changes in the rumen fermentation of sheep compared to the trehalose non-producing strain. It effectively stabilised rumen pH, promoting microbial activity that resulted in a higher SCFA concentration and the improved ruminal degradation of starch and NDF. The higher trehalose content apparently increased the viability of cells in the dried state. The effect of yeast supplementation appears to be influenced by the composition of the diet also. In case of a high-concentrate diet, the positive impact of yeasts on rumen fermentation may be more pronounced.

Due to the experimental design and the low number of samples, general conclusions cannot be drawn. It must be taken into consideration that the same animals were fed with different feeds subsequently. But the results are promising, and based on this pilot study, a larger scale experiment can be designed to be able to generalise the results and to measure the effects more precisely.

The results suggest that the T2 supplementation can be recommended in the feeding of ruminants, for example, in the early phase of lactation in dairy cows or dairy sheep. The improved NDF degradability could increase the energy supply of animals. A higher molar proportion of propionate, a gluconeogenetic substance, promotes the glucose availability.

Funding

Our research was supported by the National Office for Research and Technology [grant number OMFB-1213/2004].

Additional information

Funding

Funding: Our research was supported by the National Office for Research and Technology [grant number OMFB-1213/2004].

Notes

This article was originally published with errors. This version has been corrected. Please see Erratum (http://dx.doi.org/10.1080/09712119.2013.889900)