Effect of sodium formate and lactic acid bacteria treated rye silage on methane yield and energy balance in Hanwoo steers

Silage preparation

The forage rye was harvested using a rotary mower (Frontier RC2048, John Deere, IL, USA) in the morning of May 2019 and ensiled by noon on the same day using a non-chopper round baler (460M, John Deere) with no additives (control), either with LAB inoculant or Na-Fa (Fig. 1). The LAB inoculant (Lactobacillus plantarum 1.5 × 10¹⁰ CFU/g fresh matter; CM Bio, Anseong, South Korea) was diluted to a concentration of g/L of tap water and sprayed on the forage rye during the wrapping at a rate of approximately 4 L/ton of fresh forage rye. Sixteen percent of the Na-Fa solution was sprayed at a rate of approximately 6.6 L/ton. The optimum baler running speed was determined on the pre-manufacturing day to ensure the planned spraying amount of the additive solution. The finished bales were wrapped in 15 layers of film and stacked for 60 days at a farm at Seoul National University (latitude: 37.5202213; longitude: 128.4439284; 1447-1, Pyeongchang-daero, Pyeongchang-gun, South Korea). Prior to bale silage preparation for each treatment, forage rye was collected at approximately 10 m intervals for chemical analysis. More than four bale silages were selected from each treatment after 8 weeks of fermentation, and nine core samples were obtained from each bale silage using a sampler (Push Type t-handle Uni-forage Sampler, Star Quality Samplers Inc., AB, Canada). The core samples were pooled to produce one sample per bale, and the samples were frozen until analysis of fermentation characteristics and chemical composition.

Animals, experimental design, and diet

The experimental animals (Hanwoo) were obtained from commercial cattle auctions in South Korea. The cattle were housed in feedlots (W × L, 5 × 10 m) per four cattle at the Seoul National University farm (latitude, 37.5202213; longitude, 128.4439284; 1447-1, Pyeongchang-daero, Pyeongchang-gun, South Korea) until experimentation. The experimental feed (Table 1) was provided once daily at 10:00, and water was provided ad libitum. The feed consisted of Timothy hay and commercial concentrate feed until the end of the experiment. The cattle were raised on the same farm until slaughter at the end of the experiment. In vivo experiments were conducted with the approval of the Seoul National University Institutional Animal Care and Use Committee (approval number: SNU-181127-12).

Growing Hanwoo steers groups 1 (average body weight 275 ± 8.4 kg, n = 3) and 2 (average body weight 360 ± 32.1 kg, n =3) were randomly allocated into two adjacent pens with a duplicated 3 × 3 Latin square design, and the experimental unit was each animal. In each pen equipped with individual feeding gates, three steers individually consumed 2.7% of their body weight on a feed basis of different silages (control, LAB, and Na-Fa) and 0.2% of concentrate in the morning (09:00) and evening (18:00). The ingredient composition of the concentrate was DM = 3.19%, crude protein (CP) = 16.69% of DM, EE = 5.0% of DM, ash = 6.93% of DM, and crude fiber = 4.09% of DM. Bale silage was chopped using a TMR mixer for 5 min, packed into four or five bags, and stored at room temperature. Each period comprised 10 days of adaptation to the pen and 9 days of measurement in a direct respiratory chamber. Steers had free access to water and mineral blocks throughout the experiment. The body weights of the steers were measured at the beginning and end of the experiment. On the same day, when group 1 returned to their pen after completing the measurements in the chamber, group 2 moved to the chambers because the steers had already completed feed adaptation in the pen. After 1 day of adaptation in a chamber, feces and urine were collected for 5 days, CH₄ production was measured for 2 days, and rumen fluid was collected on day 19. All in vivo experiments were performed in accordance with IPCC guidelines.

Sampling and data collection

The silage and concentrates offered for each treatment were sampled, bulked for each period, and stored in a cool room for DM measurement and composition analysis. The total weight of feces excreted daily on a polyvinyl chloride sheet spread on the floor of each chamber was collected and weighed at 09:00 daily for 5 days for each period. Ten percent of homogenized daily feces from each animal was sampled and stored at −20 °C until the formation of composite samples by the steer within each period to calculate DM digestibility (DMD), organic matter (OM) digestibility, CP digestibility, and NDF digestibility. Urine was separated using a rubber pyramidal funnel to prevent urine contamination of the feces. When the urine was excreted, it was collected in a container, because the funnel was connected to a rotary vane vacuum pump using polypropylene tubing. Rumen fluid was collected from the rumen before morning feeding and 1.5 and 3.0 h after feeding and squeezed through four layers of cheesecloth to measure ruminal pH and analyze VFA and ammonia nitrogen (NH₃-N) concentrations.

CH₄ sampling methods and respiratory-metabolic chamber

CH₄ production from enteric fermentation by steers fed the three silages was measured using three indirect open-circuit respiratory chambers. Each chamber (Fig. 2, outside dimension: 137 cm wide × 356 cm deep × 200 cm tall) made of steel frames had a feeder and a water bowl and maintained 18–25 °C and 50–60% relative humidity by an independent regulatory system that contained an air conditioner (Dryer DK-150E, Hangzhou Duokai Technology Co., Hangzhou, China) with a recirculating fan and air filter (model ALFFIZ-WBCAI-015H, Busung, India). For recirculation, air exited the chamber (0.6 m/s) through openings in the middle of the chamber ceiling, and the dehumidified air was recycled into the chamber (0.1 m/s) through openings at the back end of the chamber ceiling. The air circulation speed was fixed to avoid positive pressure inside the chamber and to achieve maximum gas recovery. The condensed water from the refrigeration unit was drained into the chamber through a capillary tube attached to the water traps.

Each chamber was maintained at a negative pressure using a sealed rotary exhaust pump connected to a flow meter (model LS-3D, Teledyne Technologies, Thousand Oaks, CA, USA) that provided a constant wet ventilation rate of 450 L/min. The air outflow from each chamber was subsampled from the PVC pipe between the flow meter and exhaust pump using the gas sampling pump of the analysis system. All three chambers shared a common gas analysis system that comprised a multiplexer (Metabolic controller, B.S. Technolab, Seoul, South Korea) connected to a gas sampling pump (Columbus Instruments, Columbus, OH, USA) and analyzers (model VA-3000, Horiba Ltd., Kyoto, Japan) containing nondispersive infrared CH₄ (0–2,000 ppm). The air was subsampled every 14 min with a sequence of background air and chamber air sampling of 90 and 120 s for each chamber. The average value at the final sampling time of 20 s was used for the calculations. The recovery rate in each chamber was tested at the beginning of the experiment using a standard CH₄ gas mixture (25% mol/mol balance N₂; Air Korea, Seoul, South Korea). CH₄ sampling was performed according to the IPCC guidelines.

Chemical composition analysis

Fecal samples were dried for 72 h in a convection dry oven (model VS-1202D9, Vision Scientific, New Milford, CT, USA) at 65 °C, and the DM content was calculated. The dried samples were ground to pass through a 1 mm sieve (Thomas Scientific Model 4) and used to analyze the nitrogen (AOAC method 990.0), crude ash (AOAC method 942.05), and EE (AOAC method 920.39) contents. NDF and ADF contents were estimated using the methods of Van Soest, Robertson & Lewis (1991) and Van Soest (1973), respectively.

Rumen gastric juice was pretreated using the method described by Erwin, Marco & Emery (1961) for VFA analysis. A gas chromatograph equipped with a flame ionization detector and an FFAP CB column (25 m × 0.32 mm, 0.3 μm; Agilent Technologies, Santa Clara, CA, USA) was used to measure the VFA contents. The NH₃-N concentration was measured using a modified colorimetric method (Chaney & Marbach, 1962).

Rumen fermentation characteristics

The rumen fermentation characteristics of rye silage, LAB-treated rye silage, and Na-Fa-treated rye silage in Hanwoo cattle are shown in Table 4. The pH and total VFA content of the rumen fluid did not differ among the treatment groups. The ratio of acetic acid in the control, LAB, and Na-Fa groups were 69.9, 67.5, and 68.5% mol, respectively (p = 0.044). The ratio of propionic acid in the control, LAB, and Na-Fa groups were 17.0, 19.6, and 18.4% mol, respectively (p = 0.017). Butyric, iso-butyric, and valeric acid contents of the rumen fluid did not differ among the treatment groups. The ratios of isovaleric acid in the control, LAB, and Na-Fa groups were 1.23, 1.39 and 1.58% mol, respectively (p = 0.004). The NH₃-N content of the rumen fluid did not differ among the treatment groups.

Rumen contains a complex microbiota comprising bacteria, ciliated protozoa, archaeal fungi, and viruses that degrade a broad range of feed ingredients (Firkins & Yu, 2015). A large population of microbiota in the rumen protects against various shocks caused by feed ingredients. However, when large amounts of rapidly degradable ingredients are supplied to the rumen, its pH decreases, leading to a condition known as subacute ruminal acidosis (SARA) (Kleen et al., 2003). SARA is typically diagnosed when pH drops below 5.5 (Kleen et al., 2003). This study showed pH values of 6.99–7.06, and therefore, it is considered that the silage had little effect on decreases in pH and did not lead to SARA. In the present study, the control group had the highest levels of acetic acid in the rumen (Table 4). This difference in acetic acid production could be attributed to variations in the acetic acid and NDF contents of the experimental feed, because the production of acetic acid in the rumen is dependent on the NDF content of the forage (Sutton et al., 2003). In the rumen, propionate is primarily produced by propionic acid bacteria such as Selenomonas, Ruminobacter, Prevotella, and Clostridium through the succinic pathway. This pathway involves key enzymes such as propionyl CoA carboxylase and succinate CoA synthetase (Wang et al., 2020). The succinate pathway is initiated by providing pyruvate (a three-carbon molecule) derived from glucose (a six-carbon sugar). Hexose in cattle feed contains WSC and non-fiber carbohydrate fractions (Yang, 2005). Lactic acid is an important precursor of VFAs in the rumen, and there have been reports that lactic acid changes to acetic acid and propionate approximately 40% and 32% of the time, respectively (Gill et al., 1986; Nakamura & Takahashi, 1971). In rumen, butyric acid converts 93% of acetic acid or acetic acid freely converts to butyric acid (Gill et al., 1984). In the present study, the elevated acetic acid content in the rumen of the control group could be attributed to the provision of butyric acid in the diet. Conversely, it is evident that the increased propionate content in the rumen was a result of lactic acid originating from the experimental feed.

Isoacids are synthesized mainly as metabolites by the degradation of amino acids, such as valine, isoleucine, leucine, and proline, which are converted into amino acids and branched fatty acids by rumen microbes (Andries et al., 1987). Because the CP content of the LAB and Na-FA groups in the silage was greater than that of the control group (Table 4), the production of isoacids in the rumen was considered to have been influenced. Additionally, iso-butyric and valeric acids in the LAB and Na-FA groups tended to be higher than those in the control group (Table 4), which could be attributed to the higher protein intake of the LAB and Na-FA groups compared to the control group (Table 5). In this study, LAB and acid treatment resulted in an increase in silage protein content, which is thought to elevate isoacids (iso-butyric and iso-valeric acids) in rumen fluid when fed to ruminants. Therefore, feeding silage to ruminants may increase the amount of isoacids in the rumen, which are specific nutrients for ruminal cellulolytic bacteria, and appear to positively affect rumen microbial fermentation (Fievez et al., 2012). Consequently, feeding silage to ruminants might result in a weak increase in nutrient digestibility in the rumen.

Nutrients digestibility

Nutrient digestibility of rye silage, LAB-treated rye silage, and Na-Fa-treated rye silage by Hanwoo cattle is shown in Table 5. The intake of DM, CP, and NDF did not differ among the treatment groups, and the digestibility of DM, OM, CP, and NDF did not differ among treatment groups.

Silage is preserved by acidification via microbial fermentation, and various microorganisms consume nutrients for microbial growth during silage production (Muck et al., 2018). It decreases the DM and OM contents of forage, and carbon sources can be converted and excreted as carbon dioxide and fatty acids. Thus, silage production can result in a decrease in nutrients in the forage based on DM. Moreover, long-term storage of silage can lead to increased costs owing to nutrient loss. Although the DM content of forage can decrease during silage production owing to the conversion of nutrients to fatty acids, fatty acids have some advantages, such as an increased digestibility rate of feed compared to non-silage forage in the rumen (Ferraretto, Shaver & Lauer, 2014). Ruminants absorb and utilize VFAs produced by microorganisms as nutrients in the rumen. VFAs produced by microbes during silage production can be utilized by being absorbed directly through the inner rumen-wall cells (Bedford et al., 2020). In this study, the digestibility of all the nutrient fractions in the experimental diets did not differ significantly among the treatments. Although feeding trials and sampling were designed to reduce experimental errors, the complex nature of ruminal gut digestion makes it difficult to fully explain this situation. Several studies have reported improvements in DM and NDF digestibility per unit of forage feed with silage (Miron et al., 2007; Cotanch et al., 2012). Although silage has been shown to improve the digestibility of DM and NDF per unit of forage feed, the DM intake of silage forage is often lower than that of non-silage forage (Charmley, 2001). This means that although the digestibility per unit of forage may be higher, the overall feed intake may be lower, potentially resulting in a lower overall nutrient intake. Therefore, it is important to carefully consider the trade-offs between increased digestibility and potentially lower feed intake when silage is used as feed for ruminants.

Energy balance

The effects of LAB and Na-Fa treatments on the energy balance of Hanwoo cattle are shown in Table 6. GE intake did not differ among the treatment groups. Energy loss from feces, urine, and CH₄ did not differ among the treatment groups. Energy loss from heat was significantly greater in the LAB group than in the Na-FA group (p = 0.045). The ratios of energy loss from feces, urine, and CH₄ did not differ among the treatment groups. The ratio of energy loss from heat was significantly lower in the Na-FA group than that in the other groups (p = 0.049). Digestible energy (DE) and metabolizable energy (ME) did not differ among the treatment groups. The net energy (NE) of the LAB-treated group was significantly higher than those of the other groups (p = 0.010). The proportion of DE and ME (% of GE basis) did not differ among the treatment groups, and NE (% of GE basis) was significantly lower in the Na-FA group than in the other groups (p = 0.009).

The GE of the feed was measured as the energy required for complete combustion with oxygen, using a bomb calorimeter (AOAC, 2005). However, this does not indicate the total energy available in the feed. Generally, in animals, a portion of the energy in feed is released in various forms, such as feces, urine, gases, and heat, during digestion (National Research Council, 2016). The ME value is determined by subtracting the urine and gas energy from the DE value; the gas energy excreted by ruminants is mainly CH₄ (National Research Council, 2016). In this study, energy loss values from feces, urine, and CH₄, including DE and ME values calculated using the energy loss values, did not differ among the treatment groups. However, the DE tended to be higher in the LAB and Na-FA group than in the control group (p = 0.083). The heat energy loss of the LAB group was significantly greater than that of the other treatment groups, and that of the Na-FA group was the lowest among all the treatment groups (Table 6, p = 0.045 and p = 0.049, respectively). The NE of the LAB group was the lowest among all treatment groups, and that of the Na-FA groups was the lowest among all treatment groups (Table 6; p = 0.010 and p = 0.009, respectively). The change in heat energy is a complex concept that is difficult to accurately determine in animals. Thus, in live animals, heat increment is generally determined by subtracting the heat loss of fasting animals from that of eating animals (Cherian, 2019). Heat increase occurs due to the act of eating, chewing, digesting feed, and absorbing nutrients from the gut in livestock animals, assuming that they live in similar environments. Thus, the difference in heat loss in this study was considered to be caused by the effect of the additives. In recent studies in which silage inoculated with LAB was fed to ruminants, changes in microbial communities were influenced by changes in rumen metabolites (Han et al., 2022). Heat production in the rumen is caused by glucose accumulation, which occurs after an increase in glucose concentration in the rumen (Russell, 1986). Lactic acid is synthesized as the end metabolite of glucose metabolism and is utilized in gluconeogenesis and oxidation (Larsen & Kristensen, 2013). Thus, heat loss in the LAB group could be attributed to an increase in heat loss due to an increase in lactic acid supply.

Formic acid is the major secondary metabolite produced by various microbes in the rumen and contributes as much to CH₄ production as acetate and hydrogen gas (Asanuma, Iwamoto & Hino, 1998). The direct addition of formic acid during silage production contributes to an increase in WSC and hydrolysis of the NDF fraction in the forage feed (Muck et al., 2018), which reduces the population of microbial communities vulnerable to acid. This could improve nutrient absorption in the rumen by supplying rapidly degradable carbohydrates instead of LAB. However, in the present study, the lactic acid content in the Na-FA group did not differ from that in the LAB group (Table 3). Even if the usage of additives was similar between a previous study (Franco et al., 2022) and this study, the effect of additives was not the same during silage production. The differences between both experiments are as follows: 1) Franco et al. (2022) used a mixed additive (formic acid, Na-Fa, propionic acid, and potassium sorbate), whereas this study used only 60% Na-Fa, and 2) the fermentation period differed (previous study: 93 days; this study: 60 days). The key difference between formic acid and Na-Fa is that the hydrogen ions are replaced by sodium ions to form salts. Formic acid is hydrophilic, readily soluble in water, and dissociates into formate ions (HCOO^-) and hydrogen ions (H⁺) (Wu et al., 2004). The formate ion passes freely through the cell walls of bacteria, destroys the DNA of microbes, and the remaining hydrogen ions lead to further acidification (Lastauskienė et al., 2014). Na-Fa is hydrophilic, readily soluble in water, and dissociates into formate ions (HCOO⁻) and sodium ions (Na⁺) in water. Na-Fa does not release any hydrogen ions and does not have an acidifying ability. However, Na-Fa can prevent pathogenic microbial growth owing to dissociated formate ions (Lastauskienė et al., 2014). In this study, there was no sustained acidification due to the absence of hydrogen ions, and the acidification of the Na-FA group was considered to be caused by lactic acid from LAB growth. Thus, the effect of Na-FA is due to the acidification of lactic acid by LAB after formate ions restrain harmful microbial growth. As microbial growth by formate ions in silage decreases, it is possible that the NE for animals in the Na-FA group could be maintained compared to the other treatment groups.

CH₄ production

The effects of LAB- and Na-Fa-treated rye silages on CH₄ production in Hanwoo cattle are shown in Table 7. Daily CH₄ production, CH₄ production per DMI, CH₄ production per DDMI, CH₄ production per NDFI, CH₄ production per DNDFI, and the CH₄ conversion factor (Y_m) did not differ among the treatment groups. CH₄ production per DDMI of the Na-FA treatment group was lower than that of the other groups (p = 0.052), and CH₄ production per DNDFI of the LAB treatment group was higher than that of the other groups (p = 0.056).

In ruminants, CH₄ is primarily produced by methanogenic bacteria that use acetic acid and hydrogen ions in the rumen. The high acetic acid, hydrogen, and formate contents of rumen fluid can stimulate the production of CH₄ using methanogenic bacteria as precursors. In this study, lactic acid was an important precursor of VFAs in the rumen, and it can change to acetic and propionic acids approximately 40% and 32% of the time, respectively (Nakamura & Takahashi, 1971; Gill et al., 1986). In this study, the acetic acid concentration was high in the control diet, and the lactic acid concentrations in the diets of the LAB and Na-FA groups were high (Table 3). In rumen, propionic acid only converts approximately 12% to acetic acid (Gill et al., 1984). This indicates that less lactic acid can be converted to acetic acid, which may produce less CH₄ by methanogenic bacteria in the rumen. Although formic acid or formate in the rumen is more readily converted to CH₄ by methanogens, formate is predominantly utilized by LAB, where it is converted to lactic acid. Therefore, it is reasonable that CH₄ production per DDMI and DNDFI of the Na-FA treatment group was lower than that of the other groups (Table 7). Furthermore, the key differences among the treatment groups were in the NDF and WSC contents. More acetic acid is produced by ruminal microorganisms when using forage feed than when using concentrate feed. High NDF contents produce more acetic acid in the rumen, and high concentrations contribute more to propionic acid production (Wang et al., 2020). In this study, the WSC fraction of the experimental feed in the Na-FA group was mainly a propionic acid precursor. Ruminal microbes use propionic acid to produce glucose via the gluconeogenic pathway. This indicated that the contribution of propionic acid to CH₄ production was low. Thus, the addition of Na-FA could change the NDF and WSC contents during silage production, and this could affect the reduction of CH₄ derived from digested DM and NDF in the experimental feed. In a previous study, formic acid-treated silage showed a dynamic change in the NDF to WSC fraction (Franco et al., 2022). It might be considered that treatment with formic acid instead of Na-Fa appeared to be more effective in reducing enteric CH₄ in the rumen.