The red amaranth seeds (Chia Tai seed company, Bangkok, Thailand) were bought from the local market. The experiment was randomly assigned in a factorial arrangement of two factors of roughage to concentrate and seven levels of RALP percentage of total substrate in a completely randomized design (CRD). There were two factors: factor A was two levels of roughage to concentrate ratio (R:C) at 60:40 and 40:60 of dietary substrate at 0.20 g, whereas factor B was level of red amaranth (Amaranthus cruentus, L.) leaf powder (RALP) supplementation at 0, 2, 4, 6, 8, 10, and 12% of total dietary substrate.

The RALP was harvested from the plant after 25 days of growth. The red amaranth was planted in the experimental farm of Khon Kaen University, Khon Kaen, Thailand. The plots (1 × 3 m²) were prepared, sowing seeds, and watered morning at 7 a.m. and afternoon at 5 p.m. It was sun-dried, chopped, and ground to achieve the 1 mm length. Standard chemical analyses were employed to analyze for dry matter (DM), organic matter (OM), CP (AOAC, 2012), neutral-detergent fiber (NDF), and acid-detergent fiber (ADF) (Van Soest et al., 1991). Additional chemical procedures on CT (Burns, 1971) and SP (Kwon et al., 2003) were used. Macro-minerals were determined using wet digestion (nitric-perchloric digestion), atomic absorption spectrophotometry (novAA® 350 AAS) for total Ca, K, Mg, Zn, and Fe (Uddin et al., 2016).

As a source of rumen inocula, two rumen-fistulated crossbreds Holstein Friesian dairy steer [75% Holstein Friesian and 25% Thai native breed, about 370 kg body weight (BW)] were used, as rumen fluid donors. The rumen fluid donors were fed with concentrate (14% DM of CP) at 0.5% of BW, and rice straw was offered ad libitum to maintain normal rumen ecology for 21 days before rumen collection. Before the morning feeding, 1,000 ml of rumen fluid was collected from each animal to mixed, filtered, and then moved to the laboratory. The medium combining of reduced medium with rumen fluids (2:1) into hot plate for mixed under stirring at 39°C, under CO₂. The ruminal fluid mixture filled 40 ml to all bottom and incubated in a water bath at 39°C. The in vitro incubation procedure was carried out according to the study of Menke et al. (1979), as modified and described by Kang et al. (2016).

The in vitro fermentation kinetics and gas production of all treatment samples were run intervally, starting from 1, 2, 4, 6, 8, 12, 24, 48, 72, to 96 h post-incubation (42 bottles: 3 bottles/treatment × 14 treatments). The 56 bottles [2 bottles/treatment × 14 treatments × 2 sampling time (4 and 8 h of incubation)] were separately prepared for pH, protozoa count, NH₃-N, and VFA analysis. Measurement of fermentation gas production was recorded at each time, pH was measured at 4 and 8 h while the fluid was collected and was then divided into two parts. The first part of rumen fluid (18 ml) was collected and kept in a plastic bottle to which 2 ml of 1 M H₂SO₄ was added to stop fermentation process for later analyses of NH₃-N by Kjeltec Auto 1030 Analyzer (AOAC, 2012), VFA (acetic; C₂, propionic; C₃, and butyric acids; C₄) using HPLC, Instruments by Water and Nova Pak model 600E; water mode l484 UV detector; column nova Pak C18; column size 3.9 mm × 300 mm; mobile phase 10 m MH₂PO₄ [pH 2.5] according to Samuel et al. (1997), and the second portion of 1 ml rumen fluid was collected and kept in a plastic bottle to which 9 ml of 10% formalin solution for measuring of protozoal population using total direct count by hemocytometer (Boeco, Hamburg, Germany) (Galyean, 1989). Methane production [126 bottles: 3 bottles/treatment × 14 treatments × 3 sampling time (4, 8, and 12 h of incubation)] was determined from samples collected starting from 4, 8, to 12 h post-incubation. The 10 ml of gas production was collected using a 10-ml syringe, injected into 25-ml airtight serum bottles closed with a rubber lid and aluminum cap, covered with parafilm, and then measured using gas chromatography (Instruments by GC-17A System, Shimadzu; TCD detector; column shin carbon; column size 3 mm × 3 mm, activated charcoal 60/80 mesh) (Sittijunda et al., 2010). The gas production kinetic was performed based on the Ørskov and McDonal (1979) model:

where a = gas production from immediately soluble fraction, b = production of gas from the insoluble fraction, c = constant gas production rate for the insoluble fraction (b), t = time for incubation, (a + b) = the potential gas production. y = gas produced at the time “t.” The in vitro DM degradability (%) was calculated at both 12 and 24 h post-incubation [56 bottles: 2 bottles/treatment × 14 treatments × 2 sampling time (12 and 24 h of incubation)].

All the obtained data were subjected to the general linear model (GLM) [SAS (Statistical Analysis System), 2013]. Differences among treatment means were compared by the Tukey's multiple comparison test (Crichton, 1999). The statistical parameters were R:C ratios, RALP levels, and R:C ratios × RALP levels interactions. Differences among statistical treatment parameters with p < 0.05 and p < 0.001 were taken as significant differences.

Table 1 presents details of feeds and their nutritive values on DM basis. Rice straw composited of 2.8% CP, 72.7% NDF, and 47.5% ADF, respectively. Concentrate was formulated and analyzed to contain 14.4% CP, 75% TDN, 28.9% NDF, and 17.2% ADF. The nutritive values of RALP were 15.8% CP, 40.3% NDF, 29.4% ADF, and 2.37% Ca, 2.33% K, and with 0.9% CT, 0.5% SP, respectively.

The gas production kinetic parameters are presented in Table 2. Ratio of roughage to concentrate and percentage of RALP had an effect on the fermentation process and subsequently on the gas production kinetics (c, a+b). The soluble fractions of gas production (a) and insoluble fraction of gas production (b) were not affected by R:C ratio, RALP, and interaction. The c was high in R:C ratio at 40:60 while there was no difference among treatment and interaction. The a+b values for both R:C ratio and RALP were significantly increased (p < 0.01), and there were no differences in the interaction. However, the cumulative production of gas was similar among treatment by R:C ratio, RALP, and interaction. As the level of RALP supplementation increased, regardless of R:C, the values were enhanced. In vitro DM degradability at both 12 and 24 h was significantly increased for both R:C and level of RALP supplementation (p < 0.05), and there was no difference in the interaction.

Table 3 presents the findings of pH, NH₃-N, and VFA production. The rumen pH slightly dropped (6.69 to 6.56) by ratio of R:C (40:60), as compared to 60:40 (p < 0.05), while in RALP groups, they were no changes and interactions. Notable changes to increase on rumen NH₃-N were obtained when R:C at 40:60 (p < 0.05) but did not affect the RALP groups, and there was no difference in the interaction. Remarkable changes on rumen total VFA and propionate production were obtained by R:C ratio at 40:60, as compared to 60:40, and the level of RALP increased for both levels of R:C ratio (p < 0.05) while there was no interaction. The acetic and butyric production were reduced by R:C ratio (40:60), as compared to 60:40 and RALP groups, and there obtains the interaction (p < 0.05), thus lowering the C₂:C₃ ratio correspondingly.

Table 4 shows in vitro protozoal population and methane production. The rumen protozoal population (×10⁵ cell/ml) and methane production (mol/L) were reduced by the R:C and RALP supplementation, and there were significant interactions for both factors (p < 0.05). The CH₄ production resulted in a remarkable decline of CH₄ for both R:C and RALP supplementation but more dramatically for R:C at 40:60 (p < 0.05), respectively.

Both roughage and concentrate sources can affect rumen fermentation characteristics when ingested. Roughage and its fibrous characteristics will stimulate rumination and fermentation by the activity of fibrolytic bacteria, while carbohydrate sources will be additionally degraded by amylolytic bacteria. It has been reported that rumen pH could be buffered to higher than 6.2 for efficient fiber degradation and to prevent incidence of sub-acute rumen acidosis (Zebeli et al., 2010). Ørskov et al. (1974) and Wanapat et al. (2014) reiterated the importance of R:C ratio on impacting the rumen pH, VFA production, and the variation in rumen microbiome especially fibrolytic bacteria group, namely, R. albus, R. flavefaciens, and Fibrobacter succinogenes, respectively (Koike and Kobayashi, 2001; Wanapat et al., 2014). Under this experiment, the R:C ratio of 40:60 was used and greatly influenced to fermentation characteristics. Supplementation of RALP additionally enhanced the production of propionic acid. Earlier work of Kang et al. (2014) stated the impact of banana flower powder which contained phytonutrients and high concentration of minerals which could lift up the rumen pH under high concentrate supplementation level in both in vitro and in vivo feeding experiments. Under this trial, RALP that consisted of high CP (15.8% CP), 0.9% CT, and 0.5% SP, along with high concentrations of macro-minerals especially Ca, K, and Mg. RALP could enhance rumen pH and fermentation.

The R:C ratio and RALP supplementation level did not significantly impact on the a, b, and c. The R:C ratio at 40:60 yielded higher c constant than the R:C at 60:40. The a+b data for both R:C ratio were significantly increased by the RALP supplementation level, and there obtained the interactions. The accumulated total production has no significant difference among treatments.

In vitro DM degradability (%) was measured at both 12 and 24 h of incubation. Higher RALP supplementation level remarkably increased the DM degradability for both the R:C and was more pronounced for R:C at 40:60. There were no significant differences for both R:C and RALP interactions of the supplementation level. This occurrence could be due to more available carbohydrate for R:C at 40:60 and from the incremental RALP supplementation with combined phytonutrients (CT and SP) which could enrich the fermentation process. Furthermore, higher concentration of minerals could help buffer pH especially for R:C at 40:60. Aslam et al. (1991) pointed out that the use of rumen buffer such as NaHCO₃ could be beneficial when more concentrate feed was fed. It was indicative that RALP supplementation could enhance the DM degradability; nevertheless, suitable level of supplementation level should be further investigated in in vivo trials.

Rumen fermentation end-products, namely, VFAs and NH₃-N concentration, were synthesized by rumen microbiome. Russell and Rychlik (2001), Wallace et al. (2015), and Huws et al. (2018) have emphasized the close relationship of rumen microbiome and their fermentation utilization efficiency. As these fermentation end-products will serve as important substrates for the host ruminants. Under this work, the rumen pH was not changed but maintained higher for R:C at 60:40, as compared to 40:60, as the roughage level could have attributed to the result. It was notable that rumen NH₃-N concentrations were higher for R:C at 40:60 and were declined when RALP supplementation level was increased for both R:C at 60:40 and 40:60, respectively, but there were no significant interactions. Total VFAs, C₂, C₃, C₄, and C₂:C₃ ratio were greatly impacted. The C₃ concentrations were clearly shown and enhanced by both R:C and RALP supplementation level, being more explicit for R:C at 40:60 and with increasing supplementation level of RALP, while C₂:C₃ were greatly lowered. The starch fed at high level in the diet appeared to improve the total VFA and C₃ and decreased C₂ and the C₂:C₃ ratio (Kim et al., 2012). Dietary source of RALP under this experiment has enormously supported the fermentation in which the beneficial outcomes were obtained. It was further speculated that level of RALP supplementation should be evaluated. Plants and fruit wastes that contain phytonutrients have shown their effects in modulating the rumen fermentation especially increased C₃ concentration and mitigated methane production. The CH₄ production reduced by inhibiter usually the hydrogen would be increased, which is rechanneled to other hydrogen sinks, such as propionic acid, which results in an increasing C₃ concentration (Bodas et al., 2012; Kim et al., 2012; Singh et al., 2018).

Fermentation gas such as CH₄ has been produced during anaerobic fermentation in the rumen. Johnson and Johnson (1995) reported the loss of metabolizable energy in the form of CH₄ up to 15% digestible energy. Furthermore, CH₄ is one of the greenhouse gases and it has a global warming potential 23 times that of CO₂. Hence, the mitigation of CH₄ via rumen fermentation has been the major concern. Dietary manipulation in the rumen has been receiving more interests (Hook et al., 2011). The protozoal population enumerated at 4 and 8 h of incubation and were reduced for both R:C ratio and by higher RALP supplementation level, and there was greatly interactive. As shown by many researchers that the reduction of protozoal population in the rumen has a direct effect on CH₄ production since some methanogens adhered on protozoa. In addition, the presence of flavonoid extract in the feeds could also attribute to the mitigation of rumen CH₄ by inhibiting cytoplasmic membrane function and cell wall synthesis contained in protozoal and methanogens (Sommai et al., 2021). As presented under this work, the rumen CH₄ production was mitigated by increasing level of RALP supplementation for both R:C ratio at 60:40 and 40:60, respectively. This result could very well-support that RALP which could be highly promising to be employed in feeding to ruminant in order to improve rumen fermentation efficiency and mitigate CH₄ production. Ampapon and Wanapat (2021) reported that supplementation of fruit peels' powder containing phytonutrients can reduce protozoa and methane production in dairy cows. Similarly, Gunun et al. (2018) reported that in vitro CH₄ production was lower in the rambutan peel supplementation.

Under this investigation, plants rich in phytonutrients and minerals such as RALP have a promising role to modulate rumen fermentation in maintaining pH, which enhances propionate production, mitigates CH₄ production, as well as increases DM degradability. In addition, the potential use of the R:C ratio at 40:60 could enhance gas production kinetic, nutrient degradability, and rumen fermentation. RALP supplementation demonstrated potential use as a rumen enhancer and deserves further in vivo trial investigation.