Influence of Pediococcus acidilactici and Bacillus coagulans on In Vitro Ruminal Greenhouse Gas Production of Fermented Devilfish in Livestock Rumen Contents

Abstract

This study aimed to evaluate the effect of including silage from devilfish waste (SF-Hypostomus plecostomus) and probiotics (PB-Pediococcus acidilactici BX-B122 and Bacillus coagulans BX-B118) in ruminants on greenhouse gas production. The diets evaluated contained 0, 8, 14 and 20% of silage made from SF and the addition of PB at a dose of 0.2 mL/g of diet, using steers and sheep (rams) as rumen inoculum donors in a completely randomized statistical design with a 2 × 4 × 2 factorial arrangement. Asymptotic gas production (GP) was influenced (p < 0.01) by the interactions between rumen liquor (RL), SF, and PB. The inclusion of SF and PB resulted in a higher (p < 0.01) GP rate in sheep; however, the values were reduced with increasing levels of SF. Asymptotic CH₄ in the rumen fluid of steers decreased with an increasing SF percentage up to 14%. Probiotics had different effects on the rumen fluid of sheep and steers. In steers, probiotics substantially reduced (p < 0.01) CH₄ synthesis while supplementation increased it in sheep rumen fluid. Similarly, diets with probiotics had higher CO formation (p < 0.05) in sheep and steer liquor. Similarly, CO decreased (p < 0.05) with increasing levels of SF. In the rumen fluid of sheep and steers, the probiotics were found to reduce H₂S, while there was an SF-dose-dependent decrease in H₂S concentration. The ruminal pH and dry matter digestibility of sheep were higher than in steers. It can be concluded that increasing SF levels generally reduced the total gas and CH₄ production, with probiotics further enhancing this reduction, especially in CH₄ per unit of gas.

1. Introduction

Livestock production is a significant contributor to anthropogenic greenhouse gas (GHG) emissions, particularly methane (CH₄) from ruminants. In agriculture, the livestock sub-sector is reported to contribute the majority of global GHG emissions, primarily through enteric fermentation, excretions, and respiration, resulting in the release of CO₂, CH₄, and N₂O [1,2,3]. According to Bruinsma [4], global methane emissions from livestock production could increase by 6% by 2030, assuming a consistent linear relationship between CH₄ emissions and livestock population growth. While these gases are classified as short-lived climate forcers [5] and biogenic in origin, their increasing atmospheric concentrations, driven by rising demand for meat and milk, along with the global push toward net-zero livestock production, highlight the urgent need to identify effective feed additives to reduce emissions, particularly ruminal CH₄. In addition to greenhouse gas mitigation, there is a need to find an alternative source of cheap feed resources. The high cost of feed and the need for feed sustainability suggest that localized solutions may be more effective [6]. These could come from cheap resources that might be inedible or waste.

Plecostomus species (Family: Loricariidae), commonly known as toadfish or devilfish (DF), is an invasive species that appeared in Mexican freshwater systems in the 1990s (see Figure 1). It has since become a significant environmental problem for small-scale fishermen in Guerrero State and other parts of Mexico [7,8]. Invasive species are notoriously difficult to manage, raising the question of whether devilfish (DF) can be commercially exploited to offset the revenue losses faced by small-scale fishers. One potential solution is to produce fish meal from DF or utilize DF waste. DF could serve as a high-protein supplement for ruminant animals, particularly in the form of silage [8,9].

Biological silage refers to silage prepared through the addition of natural microorganisms to improve outcomes such as nutrient preservation, palatability, and digestibility. According to Salas et al. [10], DF silage contains (%): 64 dry matter (DM), 23.3 ash, 50.2 protein, 21.8 fat, 0.1 fiber, 4.6 nitrogen-free extract (NFE), and 39,844 kcal/kg of gross energy. As fresh fish, it contains 27.9% DM, 33.3% ash, 16.85% lipid, 1.92% fiber, 37.07% crude protein, and 4559.1 kcal/kg of gross energy [11]. Tejeda et al. [7] in lamb and Garcia et al. [12] in equine in vitro studies have shown the potential of devilfish silage. However, this research was performed in lamb and equine feces while neglecting cattle rumen fluid. Furthermore, none of these works showed the impact of devilfish on the fermentation profile and digestibility. In addition, those previous works also showed that fermentation can be affected by higher doses of devilfish silage. Therefore, there is a need to find additives that could help improve the fermentation profile of ruminant species in vitro.

It is well known that administering or supplementing probiotics in ruminant diets enhances productivity by stabilizing the ruminal microbiota [13,14,15]. According to Copani [16], Pediococcus acidilactici is a promising candidate for use as a direct-fed microbial to enhance the health and performance of ruminant animals. Similarly, Bacillus coagulans has also been previously administered to ruminants [17]. P. acidilactici and B. coagulans are both Gram-positive, lactic-acid-producing bacteria and thrive in acidic environments. In addition, B. coagulans is capable of producing antimicrobial compounds (like bacteriocins). Ensiling devilfish (Hypostomus plecostomus) is a strategic approach to optimizing the fermentation process and improving the nutritional quality of the resulting silage. This method can help to balance the carbon and nitrogen content. Molasses is rich in fermentable sugars, which serve as a readily available energy source for lactic acid bacteria [18]. The need to combine ensiled fish with probiotics is supported by the study of Asnawi et al. [11], where cysteine and histidine were not detected, suggesting that essential amino acids in the fish might be incomplete. Therefore, the objective of the present study was to evaluate the effect of including silage made from devilfish (H. plecostomus), along with probiotics (P. acidilactici BX-B122 and B. coagulans BX-B118), in a ruminant diet as an alternative means of utilizing invasive species, and its impact on total gas production, greenhouse gas production (CO and CH₄) and fermentation characteristics (pH and digestibility), using rumen fluid from steers and rams as inoculum. Another aim was to compare how each of these ruminant species will fare under the same feed. To achieve this objective, we used the rumen fluid of small and large ruminant species as inoculants for different levels of silage samples with or without probiotics (P. acidilactici BX-B122 and B. coagulans BX-B118) in a 2 × 4 × 2 factorial system.

2. Materials and Methods

2.1. Experimental Diets

Fresh live devilfish for the experiments was sourced from the municipality of Tuxpan lagoon, Iguala Guerrero, Mexico. The collected fish was cleansed with water to strip off solid contamination such as soil particles. The ensiling of devilfish waste was carried out in triplicate, and after milling, 5 kg of devilfish waste was combined with molasses (5%) and natural yogurt (5%, Yoplait natural^®, Guerrero, Mexico) in airtight lid bags of 20 L volume. Storage in these bags avoided leaks and air ingression, and the hermetic sealing allowed for the simple generation of an anaerobic environment. During the subsequent fermentation process, the samples were kept at ambient temperature (24–30 °C) in an area without direct sunlight and without excessive humidity (60–80% relative humidity (RH)). After 30 days, fermentation was stopped, the bags were opened, and fresh silage was taken to make the diets by mixing with the ingredients as described in detail in

Table 1. Chemical composition of pre-ensiled molasses, yogurt, devilfish waste (DF, (Hypostomus plecostomus)), and the mixed DF–molasses–yogurt silage.

Ingredient	%
Molasses
Sucrose	320
Water	18
Glucose	10
Ash	12
Potassium	7
Calcium	8
Magnesium	6
Sodium	3
Protein	2
Sulphates	6
Amino acids	0.7
Non-nitrogenous acids	3
Wax, sterols, and phosphatides	0.1
Biotin, ppm	0.9
Riboflavin, ppm	4.0
Sucrose	27
Yogurt
Total proteins	3.4
Total fats	1.8
Saturated fats	1.1
Trans fats	0
Carbohydrates	12.8
Total sugars	11.5
Added sugars	7
Dietary fiber	0
Na (mg)	50
Ca (mg)	230
Energy content (kcal)	800
Devilfish waste (DF)
Dry matter	9.54
Crude protein	31.91
Ether extract	2.93
Ash	26.75
Organic matter	73.25
Mixed DF–molasses–yogurt silage
Dry matter	8.47
Crude protein	16.29
Ether extract	4.97
Ash	15.95
Organic matter	84.05
ME (MJ/kg DM)	13.36

and

Table 2. Ingredients and chemical composition of diets for ruminants with the inclusion of different percentages of biological silage from devilfish waste (Hypostomus plecostomus).

	Level of Devilfish Silage, %
	0	8	14	20
Ingredients, g/kg diet
Maize grain	735	675	615	555
Maize stubble	150	150	150	150
Soybean grain	90	70	70	70
Devilfish silage 1	0	80	140	200
Mineral salts 2	25	25	25	25
Chemical composition, g/kg Dry matter
Organic matter	910	910	910	905
Crude protein	120	140	140	140
Neutral detergent fiber	400	420	420	520
Acid detergent fiber	220	220	230	350
Metabolizable energy	11.95	11.96	11.95	12.97

¹ Silage made with devilfish waste; the pH at opening was 4.0. ² Cu: 21.18 ppm, Fe: 4971.66 ppm, Zn: 343.75 ppm, Ca: 9.96%, Mg: 0.2495%, K: 0.8895%, Na: 1.296%, Pb: 0.0029%, P: 14.395%, S: 3.125%.

. Four formulated diets for ruminant livestock were evaluated, with an inclusion of 0, 8, 14, and 20% of biological silage made from devilfish (Hypostomus plecostomus) waste and the addition of a probiotic based on Pediococcus acidilactici BX-B122 (1 × 10¹¹ cfu/mL) and Bacillus coagulans BX-B118 (1 × 10¹¹ cfu/mL) at a dose of 0.2 mL/g of diet (cfu = colony-forming units). The ingredients for the diet and devilfish waste (viscera, skeleton, and head) for the silage were obtained from a local feed store and fishmonger in the municipality of Iguala de la Independencia, Mexico (Figure 2).

2.2. Chemical Composition of Silage and Diets

When the silages were opened, a 25 g sample from each setting was collected, and the pH was measured following the methodology described by Cherney and Cherney [19] to confirm proper fermentation. Additionally, three representative samples were taken from each diet, dehydrated at 60 °C for 72 h, and ground using a hammer mill (Laboratory Mill “model 4,” Thomas Scientific™, Swedesboro, NJ, USA) equipped with a 1 mm mesh, as described by Elghandour et al. [20].

The ash content (g/kg DM) was determined using method ID 942.05, and the nitrogen (N) content was analyzed using method ID 954.01, both according to the official methods of the Association of Official Analytical Chemists [21]. Neutral detergent fiber (NDF) and acid detergent fiber (ADF) contents were quantified using the ANKOM200 Fiber Analyzer (ANKOM Technology Corp., Macedonia, NY, USA) and the methodology of Van Soest et al. [22]. Sodium sulfite and the enzyme α-amylase were used as reagents for NDF analysis, and both NDF and ADF values were reported without residual ash, as is standard practice. Organic matter (OM) and crude protein (CP) values were calculated as described by Novriadi et al. [23].

OM = 100 − ash

CP = N × 6.25

2.3. In Vitro Incubations

The ruminal content donor animals included four steers (430 ± 20 kg body weight (BW)) and four male sheep (40 ± 5 kg BW) that were slaughtered in an official slaughterhouse under the supervision of the Mexican Official Standard. Incubations were repeated in three weeks (runs), and each run was used as an experimental unit. The rumen contents were collected weekly from slaughtered animals of the same producer, same lot, or group of animals that were slaughtered in various weeks at the slaughterhouse. Donor animals were almost fed on the same diet, and were the same breed, average body weight, age, and feed that they ingested during the rumen content collection.

Rumen contents were collected immediately after slaughtering and transported to the laboratory in airtight, insulated flasks. The contents were then filtered through four layers of gauze to obtain rumen fluid, which was subsequently used as inoculum for incubation.

The nutrient medium was prepared following the methodology of Goering and Van Soest [24], using a buffer solution, macrominerals, microminerals, a reducing agent, and resazurin. Incubations were conducted in 120 mL glass vials, to which 500 mg of diet and the appropriate dose of probiotic (if applicable) were added, along with 10 mL of ruminal inoculum and 40 mL of nutrient medium. The vials were sealed with aluminum caps and butyl rubber stoppers, gently shaken, and incubated in a water bath at 39 °C for 48 h. Each incubation was carried out in three runs, with 51 vials incubated per run. Of these, 48 vials corresponded to the experimental treatments, while 3 served as blanks containing only ruminal inoculum and nutrient medium, without the diet.

2.4. Gases Measurement

Total gas production (GP) was recorded in PSI (pounds per square inch) at 2, 4, 6, 24, 26, 28, 30, and 48 h after incubation using a digital manometer with a precision of ±2% (Manometer model 407910, Extech^® Instruments, Nashua, NH, USA), following the methodology described by Theodorou et al. [25]. Methane (CH₄), carbon monoxide (CO), and hydrogen sulfide (H₂S) levels were measured using a portable gas detector connected to an external pump (Dräger X-am^®, Lübeck, Germany). The PSI values were converted to volume (mL) using a linear regression model calibrated for laboratory conditions. To prevent overestimation, accumulated gas was released from the headspace after each measurement.

At the end of the incubation, the contents of the vials were filtered using 25 µm porosity filter bags (Filter bags F57, ANKOM Technology Corp., Macedonia, NY, USA) to separate the undigested portion of the diet (residual substrate). The pH of the liquid was measured using a potentiometer equipped with a glass electrode, specifically a wireless pH electrode (HALO^®, model HI11102, Hanna^® Instruments, Woonsocket, RI, USA). The dry matter degradability (DMD, %) was calculated based on the weight of the diet residue dehydrated at 60 °C for 48 h, using the following formula:

DMD = (D_i − D_r/D_i) × 100

where D_i is the initial quantity of the diet (g); D_r is the residual quantity of diet, and 100 is the conversion factor to %.

2.5. Calculations

The NLIN procedure (NonLINear regression, which is used to fit nonlinear regression models using least squares estimation) of SAS Version 9.0 [26] was used to estimate the fractions of “b, asymptotic production”, “c, gas rate”, and “lag time” for the total gas, CH₄, CO, and H₂S according to the following model [27]:

y = b × [1 − e^{−c (t − Lag)}]

where y represents the total gas, CH₄, CO, and H₂S production at time t (hours); b denotes the asymptotic production (mL/g DM) of total gas, CH₄, CO, and H₂S; c is the production rate (mL/h) of total gas, CH₄, CO, and H₂S; and Lag is the lag phase (hours) before the production of total gas, CH₄, CO, and H₂S begins.

2.6. Statistical Analysis

A completely randomized statistical design was employed in this study, structured as a 2 × 4 × 2 factorial arrangement. Factor 1 represented the ruminal inoculum source, Factor 2 the addition of probiotics, and Factor 3 the inclusion of devilfish silage, each evaluated with three repetitions. The data of each of the 3 runs within the same sample were averaged. Mean values of each individual sample within each treatment (3 samples of each) were the experimental unit. The data analysis was conducted using the GLM procedure in SAS [26] based on the following statistical model:

Y_ijk = μ + RL_i + PB_j + SF_k + (RL × PB)_ij + (RL × SF)_ik + (PB × SF)_jk + (RL × PB
× SF)_ijk + εijk

where

Y_ijk is the response variable, Μ is the overall mean, RL_i represents the effect of the ruminal inoculum source, PB_j represents the effect of the addition of probiotics, SF_k represents the effect of the inclusion of devilfish silage, (RL × PB)_ij is the interaction effect between the ruminal inoculum source and the addition of probiotics, (RL × SF)_ik is the interaction effect between the ruminal inoculum source and the inclusion of devilfish silage, (PB × SF)_jk is the interaction effect between the addition of probiotics and the inclusion of devilfish silage, (RL × PB × SF)_ijk represents the three-way interaction effect among the ruminal inoculum source, probiotics, and devilfish silage, and Ε_ijk is the experimental error. Tukey’s post hoc test was applied for mean comparisons, with significance determined at p ≤ 0.05.

3. Results

3.1. Gas Production

Rumen fluid from steers produced higher amounts of gases than that from sheep and gas production (GP) decreased with increasing SF levels. However, more gases were produced with the inclusion of probiotics. According to

Table 3. Total gas production of diets for ruminants with the inclusion of biological silage of devilfish (Hypostomus plecostomus) without (−) or with (+) probiotic (Pediococcus acidilactici BX-B122 and Bacillus coagulans BX-B118), using ruminal inoculum of steers and sheep.

Rumen Liquor (RL)	Level of Devilfish Silage, % (SF)	Probiotic (PB)	Total Gas Production
			Parameters 1			mL Gas Total/g DM Incubated
			b	c	Lag	2 h	24 h	48 h
Steers	0	−	294.8 b	0.040 b	6.9	23.9	182.6 d	284.8 ab
		+	235.5 d	0.042 b	3.4	30.2	203.6 c	228.7 a
	8	−	299.0 b	0.039 b	5.5	26.9	199.6 bc	288.8 ab
		+	304.0 b	0.048 a	6.0	32.6	257.1 b	294.8 ab
	14	−	276.1 c	0.037 b	3.8	29.7	191.6 bc	266.2 b
		+	231.5 d	0.040 b	5.2	34.4	246.5 b	278.0 ab
	20	−	298.4 ab	0.036 ab	3.4	33.8	203.6 bc	286.5 ab
		+	283.6 c	0.057 a	7.3	35.6	222.9 c	262.2 c
		SEM 2	20.71	0.0016	2.19	0.71	8.75	9.46
		SF	0.0036	0.0092	0.0328	<0.0001	0.0070	0.0166
		Linear	0.0008	0.0152	0.9357	<0.0001	0.0355	0.0813
		Quadratic	0.3782	0.7143	0.0709	0.0824	0.0520	0.4327
		PB	<0.0001	0.0450	0.3027	<0.0001	<0.0001	0.0326
		SF × PB	0.0168	0.0497	0.1263	0.0266	0.0707	0.0088
Sheep	0	−	229.9 d	0.038 b	4.7	25.4	162.1 d	223.4 c
		+	326.3 a	0.046 ab	5.7	37.6	273.6 a	318.0 a
	8	−	177.9 e	0.016 c	7.6	28.0	75.9 f	147.4 d
		+	285.6 d	0.041 ab	3.9	38.5	239.3 ab	277.7 ab
	14	−	155.0 f	0.017 c	10.4	27.8	80.1 f	133.0 e
		+	224.1 d	0.038 ab	2.2	34.4	174.7 d	212.2 c
	20	−	197.1 e	0.016 c	6.8	29.3	80.7 f	157.1 d
		+	122.1 g	0.029 c	7.8	32.9	101.6 e	116.8 e
		SEM 2	9.56	0.003	0.94	1.58	11.42	17.71
		SF	<0.0001	<0.0001	0.3159	0.4879	<0.0001	<0.0001
		Linear	0.1559	0.0018	0.5903	0.8088	<0.0001	<0.0001
		Quadratic	<0.0001	<0.0001	0.6689	0.8937	0.0145	0.0587
		PB	0.0192	<0.0001	0.6175	<0.0001	<0.0001	<0.0001
		SF × PB	0.0233	<0.0001	0.1850	0.0626	0.0001	0.0012
Pooled SEM 2			16.13	0.0027	1.69	1.22	10.13	14.20
p-value
RL			<0.0001	<0.0001	0.2699	0.1723	<0.0001	<0.0001
SF			<0.0001	<0.0001	0.5909	0.0022	<0.0001	<0.0001
Linear			0.0042	0.0003	0.6393	0.0002	<0.0001	<0.0001
Quadratic			<0.0001	<0.0001	0.2546	0.5267	0.3758	0.1663
PB			0.9328	0.1408	0.3819	<0.0001	<0.0001	0.0012
RL × SF			<0.0001	0.0002	0.0644	0.0001	<0.0001	<0.0001
RL × PB			<0.0001	<0.0001	0.9601	0.0061	<0.0001	<0.0001
SF × PB			0.0404	0.0018	0.1327	0.0032	<0.0001	0.0001
RL × SF × PB			0.0041	0.0062	0.1775	0.5735	0.0036	0.0012

¹ b is the asymptotic gas total production (mL/g DM); c is the rate of total gas production (mL/h); Lag is the initial delay before gas total production begins (h). ² SEM: standard error of the mean. ^a–g Means in the same row with different superscripts significantly (p < 0.05) differ among the interaction effects of type of rumen liquor × devilfish silage percentage × presence of probiotic.

, diets incubated in steer rumen liquor had significantly higher (p < 0.05) asymptotic gas, gas production rate, and total gas (mL/g DM) at 24 and 48 h compared to those incubated with sheep rumen liquor. Increasing SF levels significantly reduced (p < 0.05) gas production parameters at both time points. The onset of gas production was delayed in diets incubated with sheep rumen liquor.

Probiotics led to an increased volume of total gas/g DM incubated. ^SF and PB interaction showed that the presence of PB led to increased GP with an increasing level of SF, except in 20% SF where GP was higher without PB. The interaction of RL × SF × PB showed that under steer rumen liquor, 8% SF with probiotics produced the highest amount of gas, while in sheep rumen fluid, 0% SF with probiotics produced the highest GP and highest amount of total gas/g DM incubated for both 24 and 48 h (Figure 3).

3.2. CH₄ Production

The rumen liquor source did not affect CH₄ production at 48 h. Methane production decreased with increasing SF levels and was lower with the inclusion of probiotics. The result showed that the rumen liquor source did not affect asymptotic CH₄ production. However, the rate of methane production was higher in sheep than in steer liquor, and the time for methane production was more prolonged (delayed) in sheep liquor than in steers. At 2 and 24 h of incubation, more methane was produced from steer rumen liquor, while in 48 h, sheep rumen liquor produced (p < 0.05) more methane than steer rumen liquor. The result showed that asymptotic CH₄, the rate of gas production per hour, lag time, and mL CH₄/100 mL gas total at 2, 24, and 48 h decreased with increasing SF. Probiotics did not affect asymptotic CH₄, the rate of CH₄ production per hour, and the time before CH₄ was produced. However, the presence of probiotics reduced (p < 0.01) the proportion of CH₄ per 100 mL of biogas in 24 h. The interaction of RL × SF × PB showed that in 24 h, 8% SF with probiotics produced the lowest (p < 0.05) proportion of CH₄ in 100 mL biogas in steers, while in sheep rumen fluid, 0% SF without PB yielded the lowest proportion of CH₄ in 100 mL biogas (

Table 4. Methane (CH₄) production of diets for ruminants with the inclusion of silage of devilfish (Hypostomus plecostomus) without (−) or with (+) probiotic (Pediococcus acidilactici BX-B122 and Bacillus coagulans BX-B118), using ruminal inoculum of steers and sheep.

Rumen Liquor (RL)	Level of Devilfish Silage, % (SF)	Probiotic (PB)	CH4 Production
			Parameters 1			mL CH4/g DM Incubated			mL CH4/100 mL Gas Total
			b	c	Lag	2 h	24 h	48 h	2 h	24 h	48 h
Steers	0	−	53.58	0.0604	15.83	0.46 c	23.84 a	53.10	1.92 a	13.02 a	18.68
		+	19.85	0.0409	16.86	0.08 f	5.60 c	18.78	0.25 b	2.69 c	8.19
	8	−	51.82	0.0558	14.48	0.61 a	25.10 a	51.12	2.25 a	12.63 a	17.96
		+	25.52	0.0328	19.95	0.00 g	4.32 c	21.17	0.00 b	1.69 d	7.44
	14	−	25.38	0.0505	14.72	0.50 b	11.26 b	25.01	1.67 a	5.88 b	9.38
		+	19.04	0.0400	12.55	0.00 g	8.68 b	18.46	0.00 b	3.56 c	6.94
	20	−	36.13	0.0519	16.24	0.34 d	13.35 b	35.63	1.00 ab	6.50 b	12.42
		+	10.56	0.0529	14.29	0.00 g	4.30 c	10.20	0.00 b	1.88 d	4.00
		SEM 2	9.238	0.018	3.069	0.034	3.191	5.363	0.110	1.613	2.069
		SF	<0.0001	0.008	0.0025	0.0086	0.1711	0.0197	0.0001	0.0980	0.0324
		Linear	<0.0001	0.0442	0.1269	0.0109	0.0833	0.0273	<0.0001	0.0372	0.0225
		Quadratic	0.0321	0.0041	0.0004	0.3197	0.5230	0.1171	0.6684	0.3655	0.1563
		PB	0.6931	0.9203	0.1189	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001
		SF × PB	0.3416	0.2549	0.2498	0.0053	0.0414	0.0881	0.0004	0.0410	0.2069
Sheep	0	−	32.02	0.0741	20.64	0.15 e	2.84 c	30.40	0.50 b	1.75 d	12.38
		+	52.09	0.1007	23.00	0.00 g	9.68 b	51.70	0.00 b	3.56 b	17.69
	8	−	16.89	0.0363	17.00	0.09 f	3.47 c	14.58	0.33 b	4.46 b	9.54
		+	50.93	0.1337	23.52	0.00 g	9.99 b	50.67	0.00 b	3.75 b	20.75
	14	−	15.00	0.0473	13.19	0.14 e	4.14 c	13.75	0.50 b	5.13 b	9.96
		+	34.75	0.0982	21.64	0.00 g	8.23 b	34.76	0.00 b	4.69 b	16.44
	20	−	20.27	0.0593	19.07	0.10 f	2.89 c	17.93	0.33 b	3.58 b	11.25
		+	18.38	0.1015	20.94	0.00 g	5.04 c	18.46	0.00 b	4.50 b	16.38
		SEM 2	5.440	0.005	1.074	0.024	1.602	8.722	0.083	0.906	3.464
		SF	0.0111	0.0059	0.1509	0.6083	0.3435	0.0875	0.5847	0.1373	0.9281
		Linear	0.8225	0.0277	0.4324	0.3411	0.1713	0.0186	0.3322	0.1460	0.7295
		Quadratic	0.005	0.0247	0.2455	0.7079	0.4500	0.4873	0.5717	0.0647	0.6887
		PB	0.3102	0.8421	0.8154	<0.0001	0.0005	0.0056	<0.0001	0.5455	0.0111
		SF × PB	0.5875	0.3106	0.4921	0.6083	0.4411	0.2785	0.5847	0.4866	0.7980
Pooled SEM 2			7.581	0.014	2.299	0.029	2.525	7.240	0.098	1.308	2.853
p-value
RL			0.9598	<0.0001	0.0008	<0.0001	<0.0001	0.9665	<0.0001	0.0037	0.0149
SF			0.0031	0.0364	0.0256	0.0138	0.0636	0.0029	0.0001	0.4095	0.1805
Linear			0.0021	0.0953	0.8210	0.0065	0.0285	0.0014	<0.0001	0.2270	0.1199
Quadratic			0.1174	0.1561	0.0140	0.2981	0.8148	0.1504	0.4913	0.8739	0.2739
PB			0.4895	0.8252	0.4479	<0.0001	0.0043	0.5543	<0.0001	<0.0001	0.7449
RL × SF			<0.0001	0.0002	0.1892	0.0241	0.4865	0.7355	0.0004	0.0194	0.7069
RL × PB			0.2766	0.8690	0.7508	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001	<0.0001
SF × PB			0.2211	0.2995	0.2978	0.0191	0.1568	0.2260	0.0008	0.0734	0.6701
RL × SF × PB			0.9839	0.2777	0.6555	0.0069	0.0150	0.1441	0.0010	0.0388	0.4663

¹ b is the asymptotic CH₄ production (mL/g DM); c is the rate CH₄ production (mL/h); Lag is the initial delay before CH₄ production begins (h). ² SEM: standard error of the mean. ^a–g Means in the same row with different superscripts significantly (p < 0.05) differ among the interaction effects of type of rumen liquor × devilfish silage percentage × presence of probiotic.

and Figure 4).

3.3. CO Production

The volume of CO was lower for sheep rumen fluid compared to that of steers. CO production decreased as the percentage of SF increased in the diet, while more CO was produced within 48 h when probiotics were used.

Table 5. Carbon monoxide (CO) production of diets for ruminants with the inclusion of silage of devilfish (Hypostomus plecostomus) without (−) or with (+) probiotic (Pediococcus acidilactici BX-B122 and Bacillus coagulans BX-B118), using ruminal inoculum of steers and sheep.

Rumen Liquor (RL)	Level of Devilfish Silage, % (SF)	Probiotic (PB)	CO Production
			Parameters 1			mL CO/g DM Incubated
			b	c	Lag	2 h	24 h	48 h
Steers	0	−	0.0183 c	0.00001	0.0017	0.00007 c	0.00843	0.01814 b
		+	0.0748 a	0.00006	0.0124	0.00042 c	0.04358	0.07280 ab
	8	−	0.0141 c	0.00001	0.0014	0.00010 c	0.00778	0.01398 b
		+	0.0784 a	0.00006	0.0143	0.00119 a	0.04309	0.07533 ab
	14	−	0.0150 c	0.00001	0.0013	0.00010 c	0.00890	0.01490 b
		+	0.0993 a	0.00007	0.0149	0.00141 a	0.05558	0.09618 ab
	20	−	0.0165 c	0.00001	0.0015	0.00009 c	0.00817	0.01626 b
		+	0.0995 a	0.00005	0.0108	0.00134 a	0.05868	0.09799 ab
		SEM 2	0.09020	0.000036	0.00183	0.000074	0.003984	0.00508
		SF	<0.0001	<0.0001	<0.0001	<0.0001	0.1324	0.0466
		Linear	<0.0001	<0.0001	<0.0001	<0.0001	0.0811	0.0357
		Quadratic	0.0008	<0.0001	<0.0001	0.0006	0.4741	0.3494
		PB	0.9445	0.1150	0.3031	<0.0001	<0.0001	<0.0001
		SF × PB	0.8969	0.0111	0.0093	<0.0001	0.1618	0.0336
Sheep	0	−	0.2209 b	0.00005	0.0112	0.00053 c	0.14134	0.21699 ab
		+	1.4006 a	0.00009	0.0203	0.00013 c	0.53225	1.36220 a
	8	−	0.5294 b	0.00003	0.0149	0.00083 a	0.13430	0.45366 ab
		+	0.8927 a	0.00020	0.0227	0.00012 c	0.22860	0.87301 a
	14	−	0.3426 b	0.00003	0.0099	0.00120 a	0.13970	0.31584 ab
		+	0.9287 a	0.00008	0.0192	0.00010 c	0.35595	0.91568 a
	20	−	0.3447 b	0.00004	0.0205	0.00086 a	0.04616	0.27726 ab
		+	0.4585 b	0.00009	0.0198	0.00008 c	0.17935	0.44540 ab
		SEM 2	0.00529	0.000004	0.00077	0.000227	0.056396	0.080357
		SF	<0.0001	0.0262	0.0011	0.5903	0.0073	0.0006
		Linear	0.0028	0.243	0.003	0.5484	0.0011	<0.0001
		Quadratic	<0.0001	0.0179	0.0047	0.2251	0.6433	0.5706
		PB	0.0187	0.3071	0.0042	0.0003	<0.0001	<0.0001
		SF × PB	0.0008	0.2551	0.0718	0.5215	0.0785	0.0001
Pooled SEM 2			0.06389	0.000025	0.00141	0.000169	0.039977	0.056935
p-value
RL			<0.0001	0.0014	<0.0001	0.2091	<0.0001	<0.0001
SF			<0.0001	0.0002	<0.0001	0.0117	0.0037	0.0001
Linear			<0.0001	0.0157	<0.0001	0.0161	0.0006	<0.0001
Quadratic			<0.0001	0.0018	<0.0001	0.0170	0.6050	0.5275
PB			0.0137	0.3823	0.0120	0.1415	<0.0001	<0.0001
RL × SF			<0.0001	0.2827	0.0043	0.5604	0.0024	<0.0001
RL × PB			0.0135	0.2337	0.0014	<0.0001	0.0002	<0.0001
SF × PB			0.0001	0.1653	0.0118	0.7110	0.0685	<0.0001
RL × SF × PB			0.0001	0.3110	0.1162	0.0108	0.0536	<0.0001

¹ b is the asymptotic CO production (mL/g DM); c is the rate CO production (mL/h); Lag is the initial delay before CO production begins (h). ² SEM: standard error of the mean. ^a–c Means in the same row with different superscripts significantly (p < 0.05) differ among the interaction effects of type of rumen liquor × devilfish silage percentage × presence of probiotic.

shows that the rumen liquor source affected CO production. One can see that the asymptotic CO, rate of CO production per hour, lag time, and mL CO/g DM incubated for 24 and 48 h was higher (p < 0.05) in the rumen liquor sourced from sheep than that from steers. Similarly, the asymptotic CO, rate of CO production per hour, lag time, and mL CO/g DM incubated for 24 and 48 h reduced with an increasing percentage of SF. Similarly, the addition of probiotics led to an increased (p < 0.05) asymptotic CO, rate of CO production per hour, lag time (time before its initial production), and proportion of CO per gram of dry matter incubated for 24 and 48 h. The interaction of SF and PB showed that in the presence of probiotics, CO increased for each percentage of SF used. The interaction of RL × SF × PB showed that in steers’ rumen liquor, the diet containing 20% SF with probiotics produced the highest (p < 0.05) asymptotic CO, rate of CO production, lag time, and proportion of CO per gram of DM incubated for 24 and 48 h, while 8% without probiotics did the opposite. In sheep rumen fluid, 0% SF without probiotics produced the lowest (p < 0.05) asymptotic CO and the proportion of CO per gram of dry matter after incubation for 24 and 48 h (Table 5 and Figure 5).

3.4. H₂S Production

The volume of H₂S produced by the diets incubated in sheep rumen liquor was lower than that from steers. H₂S production decreased with increasing levels of SF, and more H₂S was produced in the absence of probiotics.

Table 6. Hydrogen sulfide (H₂S) production of diets for ruminants with the inclusion of silage of devilfish (Hypostomus plecostomus) without (−) or with (+) probiotic (Pediococcus acidilactici BX-B122 and Bacillus coagulans BX-B118), using ruminal inoculum of steers and sheep.

Rumen Liquor (RL)	Level of Devilfish Silage, % (SF)	Probiotic (PB)	H2S Production
			Parameters 1			mL H2S/g DM Incubated
			b	c	Lag	2 h	24 h	48 h
Steers	0	−	0.0945	0.00010 a	0.0228	0.00010	0.01615 a	0.09454
		+	0.0601	0.00005 b	0.0209	0.00000	0.00135 b	0.00603
	8	−	0.0481	0.00004 b	0.0163	0.00013	0.01390 a	0.05698
		+	0.0733	0.00005 b	0.0146	0.00002	0.00123 b	0.00656
	14	−	0.0060	0.00001 c	0.0021	0.00018	0.01613 a	0.04569
		+	0.0066	0.00001 c	0.0022	0.00002	0.00539 b	0.01102
	20	−	0.0112	0.00000 d	0.0014	0.00019	0.03313 a	0.07238
		+	0.0044	0.00001 c	0.0022	0.00002	0.00100 b	0.00437
		SEM 2	0.00287	0.000003	0.00040	0.000027	0.004645	0.008641
		SF	<0.0001	<0.0001	<0.0001	0.2401	0.2160	0.0963
		Linear	<0.0001	<0.0001	<0.0001	0.0741	0.0925	0.1870
		Quadratic	0.0001	<0.0001	<0.0001	0.3687	0.6012	0.0486
		PB	0.5234	0.0628	0.2975	<0.0001	<0.0001	<0.0001
		SF × PB	0.0202	<0.0001	0.3860	0.5259	0.1228	0.0371
Sheep	0	−	0.0521	0.00006 b	0.0150	0.00010	0.02794 a	0.05227
		+	0.0048	0.00002 c	0.0026	0.00000	0.00031 b	0.00477
	8	−	0.0364	0.00004 b	0.0141	0.00011	0.01309 a	0.03449
		+	0.0031	0.00001 c	0.0026	0.00000	0.00025 b	0.00313
	14	−	0.0320	0.00004 b	0.0130	0.00011	0.01381 a	0.03112
		+	0.0025	0.00001 c	0.0027	0.00000	0.00016 b	0.00240
	20	−	0.0392	0.00004 b	0.0142	0.00012	0.01392 a	0.03677
		+	0.0016	0.00001 c	0.0026	0.00000	0.00010 b	0.00139
		SEM 2	0.00833	0.000005	0.00087	0.000006	0.000446	0.002686
		SF	<0.0001	<0.0001	<0.0001	0.6171	<0.0001	0.0021
		Linear	<0.0001	<0.0001	<0.0001	0.2027	<0.0001	0.0029
		Quadratic	<0.0001	<0.0001	<0.0001	0.8693	<0.0001	0.0080
		PB	0.1898	<0.0001	0.8954	<0.0001	<0.0001	<0.0001
		SF × PB	0.0089	0.0010	0.1235	0.6171	<0.0001	0.0145
Pooled SEM 2			0.0062	0.00001	0.0007	0.00002	0.0033	0.0064
p-value
RL			<0.0001	0.0045	<0.0001	0.0128	0.1661	<0.0001
SF			<0.0001	<0.0001	<0.0001	0.1536	0.1457	0.0043
Linear			<0.0001	<0.0001	<0.0001	0.0389	0.7993	0.0245
Quadratic			<0.0001	<0.0001	<0.0001	0.3547	0.1661	0.0061
PB			0.2955	0.0001	0.3621	<0.0001	<0.0001	<0.0001
RL × SF			0.0031	<0.0001	<0.0001	0.3487	0.0201	0.6283
RL × PB			0.8662	0.3892	0.3083	0.3061	0.8573	0.0005
SF × PB			0.0007	<0.0001	0.3470	0.4168	0.0508	0.0024
RL × SF × PB			0.0002	0.0030	0.2566	0.6478	0.0179	0.2468

¹ b is the asymptotic H₂S production (mL/g DM); c is the rate H₂S production (mL/h); Lag is the initial delay before H₂S production begins (h). ² SEM: standard error of the mean. ^a–d Means in the same row with different superscripts significantly (p < 0.05) differ among the interaction effects of type of rumen liquor × devilfish silage percentage × presence of probiotic.

showed that the asymptotic H₂S, rate of H₂S production per hour, lag time, and mL H₂S/g DM incubated for 24 and 48 h were higher (p < 0.05) in the rumen liquor sourced from steers than that from sheep. The influence of SF showed that the asymptotic H₂S, rate of H₂S production per hour, lag time, and mL H₂S/g DM incubated at 48 h were reduced with an increasing percentage of SF. Probiotics did not affect the asymptotic H₂S, but reduced (p < 0.01) the volume of H₂S per gram of dry matter that was produced at 2, 24, and 48 h. The interaction of RL × SF × PB showed that in steers, a diet containing 20% SF with probiotics produced the lowest (p < 0.05) mL H₂S/g DM incubated, while 0% SF without probiotics produced the highest volume of H₂S/g DM incubated at 24 h. In sheep rumen fluid, 0% SF without probiotics produced the highest volume of H₂S/g DM incubated at 24 h, while 20% SF with probiotics produced the lowest volume of H₂S/g DM (Figure 6).

3.5. Rumen pH and Dry Matter Degradability

The results presented in

Table 7. Ruminal pH and dry matter degradability diets for ruminants with the inclusion of silage of devilfish (Hypostomus plecostomus) without (−) or with (+) probiotic (Pediococcus acidilactici BX-B122 and Bacillus coagulans BX-B118), using ruminal inoculum of steers and sheep.

Rumen Liquor (RL)	Level of Devilfish Silage, % (SF)	Probiotic (PB)	pH 1	DMD 1, %
Steers	0	−	6.57	83.79 a
		+	5.91	66.57 b
	8	−	6.99	85.38 a
		+	5.90	62.10 b
	14	−	6.88	85.42 a
		+	5.85	53.50 b
	20	−	7.07	86.13 a
		+	5.98	55.86 b
		SEM 2	0.072	1.933
		SF	0.0078	0.0370
		Linear	0.0012	0.0457
		Quadratic	0.8036	0.0456
		PB	<0.0001	<0.0001
		SF × PB	0.0243	0.0056
Sheep	0	−	6.81	83.47 a
		+	6.60	73.35 ab
	8	−	6.92	71.21 ab
		+	6.42	73.80 ab
	14	−	6.94	72.65 ab
		+	6.51	73.77 ab
	20	−	6.95	72.92 ab
		+	6.46	68.35 b
		SEM 2	0.053	1.080
		SF	0.7351	<0.0001
		Linear	0.9753	<0.0001
		Quadratic	0.6815	0.1794
		PB	<0.0001	0.0024
		SF × PB	0.0494	<0.0001
	Pooled SEM 2		0.063	1.565
	p-value
	RL		<0.0001	0.0952
	SF		0.0306	<0.0001
	Linear		0.0034	<0.0001
	Quadratic		0.9660	0.0148
	PB		<0.0001	<0.0001
	RL × SF		0.0120	0.0718
	RL × PB		<0.0001	<0.0001
	SF × PB		0.0007	0.0224
	RL × SF × PB		0.7914	<0.0001

¹ pH is ruminal pH; DMD is dry matter degradability. ² SEM: standard error of the mean. ^a,b Means in the same row with different superscripts significantly (p < 0.05) differ among the interaction effects of type of rumen liquor × devilfish silage percentage × presence of probiotic.

show that the ruminal pH in sheep liquor was higher than that of steers. The influence of SF was such that the rumen pH increased while dry matter digestibility (DMD) decreased with increasing SF. Probiotics influenced the fermentation parameters in a way that when probiotics were used, there was a lower (p < 0.05) rumen pH and lower DMD. The interaction of RL × SF and PB showed that in steers, diets with 20% SF without probiotics had the highest DMD, while 14% SF without probiotics had the lowest DMD. In sheep rumen fluid, the diet with 0% SF without probiotics had the highest DMD, while 20% SF with probiotics had the lowest value.

4. Discussion

4.1. Total Gas, CH₄, CO, and H₂S Production

According to Vallejo-Hernandez et al. [28], differences in the fermentation characteristics observed in different ruminant species might be due to variances in the population of bacteria and protozoa and the activities of rumen microbes. Henderson et al. [29] reported that rumen microbial diversity can be affected by factors including the diet, geographical location, feeding strategy, feed additives, and physiological condition of the host animals; therefore, the variation between the total gas produced between rumen sources of cattle and sheep may be due to the natural inherent differences between the host animals. Bao-Shan Xing et al. [30] also reported that cattle rumen liquor is highly diverse and rich in bacteria and fungi compared to sheep rumen fluid. Therefore, the reason for the higher gas produced in the rumen fluid of cattle might be influenced by the ability of their rumen microbes to adapt to biological silage due to the robustness and diversity being broader than in sheep. In addition, cow rumen microbes are generally known to secrete more extracellular multienzyme complexes to hydrolyze lignocellulosic biomass than sheep rumen microorganisms. The overall decrease in gas production (GP) with increasing SF is similar to a study reporting that asymptotic fecal gas production began to decline when SF was included up to 12 and 18% in horse diets [12]. The reason for this decline may be attributed to the imbalance in the carbon-to-nitrogen ratio where the high protein (nitrogen) content of fish silage was not matched by sufficient fermentable energy (carbon). This can limit microbial activity and reduce fermentation efficiency, leading to lower total gas production. It has been reported that protein fermentation produces less gas compared to carbohydrates [31]. Similarly, the rate of gas production, which can be an indicator of the speed of digestion, also decreased with increasing SF. This suggests that microbial fermentation was affected by the gas production. It must be remembered that the presence of lactobacilli in yoghurt during fermentation could have made use of the available carbohydrates during silage making. It is known that lactobacilli ferment milk carbohydrates (lactose) to produce lactic acid in yoghurt; thus, the lactobacilli in yoghurt could have used the carbohydrates in molasses and in turn have produced lactic acid, as reported by Astuti et al. [32], who found that lactobacilli could reduce methane; we therefore suggest that the reduction in gas production with increasing SF could be attributed to the lactic acid and lactobacilli which either reduced the ability of rumen microbes to digest the SF or that the microbes have used a lot of rapidly fermentable sugar that could have been available for rumen microbes. Looking at the CH₄ produced, it can also be seen that CH₄ reduced with increasing SF. This reduction could either be due to the ability of lactic acid to reduce methane production or simply the unavailable carbon from the SF, which would have been used to form methane. The increase in gas production in the presence of probiotic cocktails may be attributed to the ability of the probiotic cocktail to reduce pathogenic microbes and also to their carbohydrate metabolism. Chen et al. [33] reported that B. coagulans increases gas production during in vitro digestion. Therefore, the increased gas production could be due to this. The interaction of the test showed that the rumen fluid source has various influences on digestibility parameters. For example, the interaction showed that in steers, 8% SF with probiotics produced the highest volume of gases, while in sheep, 0% SF with probiotics produced the highest level of gas. The implication of this is that the microorganisms in the two rumen fluid sources reacted differently to the SF. It seems plausible that there is limited diversity and robustness in the rumen microbes of sheep, which are less adaptable or not complex enough to handle SF. Perhaps their silage metabolites could not be handled by sheep rumen microbes, because they were more sensitive to the SF. The trend showed that steer rumen microbes were more adapted to increased SF silage while sheep microbes were less adaptable, because as SF increased, gas production was reduced. To support this initial point, the overall trend showed that, except at 8% SF under steer rumen liquor, the inclusion of probiotics generally resulted in lower gas production. By contrast, under sheep rumen fluid, increasing levels of SF silage consistently reduced gas production. However, at each level of SF inclusion, the presence of probiotics enhanced gas production compared to its absence. This difference may be attributed to host-specific factors and the prior adaptation of the rumen microbes before the animals were slaughtered. Therefore, the rumen microbes from steers may have been less tolerant to the probiotics used in this study. Yet, looking at the total volume of gas produced by steer liquor, which is 29.3% higher than for sheep, it can be deduced that steer rumen microbes are still more effective at degrading the same feed sample, which is an indication of diversity, richness, and robustness. Another observation in both total gas and methane production, as influenced by the host, was the sharp increase in gas from steer rumen liquor compared to sheep liquor between 24 and 32 h of incubation in both Figure 3 and Figure 4. This may be due to physiological differences between cattle and sheep, particularly in rumen microbial composition, methanogen abundance, and fermentation capacity. Steers tend to harbor a more diverse and abundant microbial population, which contributes to more rapid fermentation and greater gas release during peak fermentation phases. Additionally, differences in rumen retention time and enzymatic activity may further explain the more pronounced gas production patterns observed in steers compared to sheep.

Methane is classified as one of the main climate pollutants. It is produced as a by-product of organic matter fermentation under anaerobic fermentation. Though biogenic, it has both metabolic and environmental costs. In this study, methane from sheep and steers’ rumen fluid was very similar, although the rate of methane production per hour was higher in sheep rumen fluid than in steers. However, looking at the proportion of methane produced for every 100 mL of biogas, cattle produced more than sheep rumen fluid at 24 h of incubation, while sheep produced more than cattle rumen fluid at 48 h of incubation. At 24 h, methane production from cattle rumen fluid was 1.52- to 4.27-times higher than that from sheep. However, by 48 h, methane production from sheep rumen fluid was 1.34-times higher than that from cattle. The reason for this may the due to the microbes in their rumen liquor. It is estimated that globally, methane emissions from sheep are at least 90% lower than from cattle, and the emission intensity of beef cattle is about 71 kg CO₂-eq/kg of carcass weight vs. 6.9 kg CO₂-eq/kg fat and protein corrected milk [34]. At 48 h, it could be that slower fermentation in sheep rumen liquor led to more methane formation than in cattle. Moreover, the decrease in methane production with increasing SF levels may be attributed to fermentation processes that produce lactic acid, which can inhibit methanogenesis. For example, Doyle et al. [35] reported that lactic acid bacteria can reduce methane production. The interaction showed that at 24 h, steer rumen fluid containing 8% SF in the presence of probiotics produced the lowest methane. This can be attributed to the methane reduction activity of probiotics, especially P. acidilactici which has been reported to reduce methane production and improve in vitro feed digestibility and fermentation [36]. P. acidilactici, like other LAB, ferment CHO to lactic acid rather than producing much hydrogen. Therefore, this reduces the hydrogen pool available for methanogens, leading to lower methane output. The lactic acid produced by P. acidilactici can be utilized by lactate-utilizing bacteria (e.g., Megasphaera elsdenii, Selenomonas ruminantium) that convert lactate to propionate.

Under sheep rumen fluid, 0% SF in the absence of probiotics produced the lowest amount of CH₄ for every 100 mL of biogas. This suggests that, in the case of sheep, neither of the additives was effective in reducing methane production. The lower methane output appears to be more influenced by the host animal itself than by the inclusion of SF or probiotics.

Carbon monoxide (CO) has been identified as an intermediate product of dry matter (DM) degradation [37], and its production is linked to incomplete feed degradation or reduced microbial activity in the rumen [38]. The production of CO by an inoculum is associated with the fermentative capacity of rumen microorganisms and their diverse activities. It is therefore one of the gases produced by the ruminal microbiota during the degradation of organic matter [39]. Consequently, the increase in dry matter digestibility observed in steers with SF inclusion was associated with reduced CO production. In contrast, the elevated CO levels in sheep rumen fluid may be linked to lower dry matter digestibility. CO gas has the potential to be oxidized to produce CO₂ [40] and can be utilized for the production of CH₄ by methanogens, as recorded in the present study. Thus, CO was observed to be an important metabolite that functions as a regulator of vital metabolic pathways like the production of CH_4. In this study, the higher CO produced, which is 12.30-fold higher than in steer fluid, could be an alternative fermentation by-product for sheep rather than CH₄. Similarly, the decreasing CO with increasing SF could be due to the ability of CO to lower the level of carbon available for microbes with increasing levels of DF silage, and the increasing nitrogen may lead to competition for carbon. The higher levels of CO observed in the presence of probiotics suggest that the methane-inhibiting effects of the probiotics led to an increased availability of CO, which was not utilized for methane formation. The interaction results showed that steers fed 20% SF with probiotics produced the highest amount of CO, while in sheep rumen fluid, the highest CO levels were observed at 0% SF with probiotics. This indicates that, in both cases, probiotics favored the accumulation of CO over its conversion to CH₄. Furthermore, the antimicrobial properties of probiotics, including their activity against methanogens [35], may have contributed to the reduction in methane and the corresponding increase in CO accumulation.

The production of H₂S in the rumen has been reported to cause toxicity and diseases arising from alterations in metabolism. Steers’ inoculum produced more H₂S than that of sheep. Alvarado-Ramirez et al. [41] also reported higher H₂S production in bovine inoculum when compared to sheep rumen fluid. Since sheep generally produce less methane than steers, and H₂ is necessary for methane formation, this suggests that the lower level of H₂S indicates that the H₂ was used for other metabolic purposes rather than methane. The lower H₂S with increasing SF silage could be due to the differences in the population of sulfur reduction bacteria, which are responsible for the reduction of sulfur to H₂S [42]. The inclusion of SF and PB could have manipulated the rumen microflora and exerted an antimicrobial effect that reduced the quantity of sulfur-reducing bacteria, thereby leading to a reduction in the production of H₂S. The reduction recorded in H₂S production confirms the assertion that the use of additives can lead to a reduction in H₂S production in the rumen [41].

4.2. Rumen pH and Dry Matter Degradability

Rumen pH is an indicator of rumen acidity or basicity. In our study, the rumen pH values of steers rumen fluid were lower than those of sheep. It was observed that DM was lower with increasing levels of SF silage. This may be due to the high nitrogen content associated with increasing SF levels, which could have caused a carbon–nitrogen imbalance and negatively affected digestion. Alternatively, the increasing levels of lactic acid resulting from the activity of lactic acid bacteria in yoghurt may have exerted antimicrobial effects on rumen microbes. The lower dry matter digestibility observed with increasing levels of SF silage suggests reduced microbial activity, indicating that less fermentation occurred. This is consistent with the known relationship between the microbial degradation of substrates, energy availability, and the production of volatile fatty acids [43]. Therefore, lower dry matter digestibility levels indicate that increasing SF silage may be toxic to the microbes or there is simply not enough feed for the microbes to act on. Another explanation might be that the metabolites in the SF silage are inhibitory to the microbes. One reason for this is the characteristics of probiotics. Pediococci is a lactic acid bacterium, and it has the ability to produce nisin (with antimicrobial activity) [44]—the use of which has been reported to increase propionate production—and glucogenic volatile fatty acids [45], while B. coagulans exhibits a versatile carbohydrate metabolism [33,46]. Therefore, probiotics have numerous enzymes for the utilization of different carbohydrates and harbor genes for antibiotic biosynthesis. Therefore, these combined activities allowed them to adapt to the rumen environment and to be more efficient in energy production despite lower digestibility. The influence of the host on the CH₄ efficiency per unit of OM fermented may be attributed to the sheep’s inherent ability to produce less methane than steers. This could reflect a lower diversity or population of methanogens, or a metabolic preference for alternative pathways that utilize carbon and hydrogen for purposes other than methane formation, as is more prevalent in sheep compared to cattle.

Summarized Discussion

The findings of this study clearly demonstrate the differential fermentation responses of sheep and steer rumen fluid to the inclusion of soybean waste (SF) and probiotics (PB). This discussion integrates the gas production profiles, methane (CH₄), carbon monoxide (CO), hydrogen sulfide (H₂S) emissions, and digestibility parameters into a cohesive interpretation of microbial dynamics and fermentation efficiency. Rumen fluid sourced from steers consistently produced higher total gas volumes and fermentation activity compared to sheep, suggesting that steer microbial populations are more diverse and robust. This aligns with earlier reports (e.g., Bao-Shan Xing et al. [30]) that cattle harbor richer microbial communities. In contrast, sheep rumen microbes appeared less adaptive to increasing SF, which may reflect either lower microbial diversity or greater sensitivity to the metabolites in SF silage. Total gas production decreased with increasing SF, likely due to the nitrogen–carbon imbalance introduced by SF, as protein fermentation yields less gas than carbohydrate fermentation. The inclusion of probiotics increased gas production overall, supporting their beneficial role in enhancing fermentation, possibly by reducing pathogens and providing additional enzymes for carbohydrate breakdown. The effect of probiotics was especially notable under steer rumen conditions with 8% SF, indicating a complex host–microbe–diet interaction. Methane production dynamics showed time-dependent and species-specific variations. Although both rumen sources produced similar overall methane quantities, their temporal patterns differed. Steers produced more methane early (24 h), while sheep surpassed them at 48 h, especially at the proportion of methane per 100 mL of gas produced. The reduction in methane with increasing SF and probiotic use is significant in light of climate change, suggesting mitigation potential via diet manipulation and probiotic supplementation. The action of lactic-acid-producing bacteria like P. acidilactici in consuming hydrogen (thereby reducing substrate availability for methanogens) may explain this effect. CO and H₂S profiles also reflected the underlying microbial processes. Higher CO levels in sheep fluid suggest incomplete fermentation or a shift in microbial pathways under less efficient digestion. The observed CO accumulation with probiotics could indicate reduced methanogenic conversion, reinforcing the methane mitigation hypothesis. The reduced H₂S production with increasing SF and PB suggests antimicrobial effects on sulfur-reducing bacteria, supporting safer ruminal fermentation. Rumen pH and dry matter digestibility (DMD) were influenced by both diet and microbial source. While sheep rumen maintained a higher pH, digestibility declined with increasing SF levels, likely due to nitrogen overload, low carbon availability, or inhibitory silage metabolites. The lower DMD with probiotics could also be linked to antimicrobial substances like nisin from Pediococcus spp., which may have altered microbial populations or reduced overall fermentation efficiency. Despite this, probiotics still improved fermentation outcomes (e.g., gas yield and methane reduction), indicating functional trade-offs.

5. Conclusions

Steer rumen fluid consistently produced more gas and fermentation end-products than sheep. Increasing SF levels generally reduced total gas and CH₄ production, with probiotics further enhancing this reduction, especially in CH₄ per unit of gas. For steers, diets with 8% SF and probiotics maximized gas and fermentation efficiency while minimizing CH₄ and H₂S. In sheep, the best fermentation profile was seen with 0% SF plus probiotics, which were delivered with lower gas emissions. Dry matter digestibility (DMD) declined with higher SF inclusion, indicating reduced nutrient breakdown at greater silage levels. Probiotics often compounded this effect, particularly in sheep, despite improving energy yield. In steers, the highest DMD was seen with 20% SF without probiotics, while the lowest was with 14% SF without probiotics. In sheep, the dry matter digestibility (DMD) was highest at 0% SF without probiotics but dropped sharply at 20% SF with probiotics, indicating a strong interaction effect between diet components and host species. These results underline the need to balance emissions reduction with digestibility. Overall, these findings suggest that moderate SF inclusion combined with probiotics can enhance fermentation efficiency while mitigating gas emissions, with diet optimization depending on the ruminant species.