Evaluation of Essential Oils as Additives during Fermentation of Feed Products: A Meta-Analysis

Abstract

Essential oils (EOs) are secondary metabolites of plants with antimicrobial functions that can be used as silage additives, but their effectiveness has been inconsistent. Therefore, this study aimed to evaluate the effects of EOs as silage additives on nutritional quality, fermentative products, microbial population, and in vitro rumen fermentation. A total of 17 articles consisting of 113 studies (comparisons) were selected for meta-analysis. The results showed that EO supplementation increased several parameters of nutrient contents such as dry matter (DM), crude protein (CP), and ether extract (EE) (p < 0.05), but decreased crude fiber content and mold population (p < 0.05). EO supplementation also affected rumen fermentability, namely reduced gas production, organic matter digestibility, and some fermentation products such as total VFA, iso-butyrate, iso-valerate, acetic, and succinic acid (p < 0.05) but did not affect methane production. Sub-group analysis based on the source of EOs revealed that only cumin and oregano reduced mold population during ensiling (p < 0.05). These results showed that EO supplementation improved several nutrient qualities such as DM, CP, and EE, inhibiting mold growth and stabilizing rumen pH.

1. Introduction

The making of fermented feed or silage is a method to maintain forage availability in the dry season. However, during the ensiling process, there is often a decrease in the physical and chemical quality of the resulting forage. To prevent this decrease in the quality of forage during storage, several approaches such as the use of additives during ensiling can be implemented. The use of feed additives is carried out in small quantities to enhance feed utilization efficiency and maintain the production and health of livestock. Previous investigations have established the use of supplementary feed in the form of antibiotics in livestock businesses to increase the efficiency of feed use. However, the use of antibiotics as feed additives has been officially prohibited because their residues can be deposited in livestock products, which endangers the health of consumers. Therefore, it is recommended to use natural feed additives for livestock, which have been widely studied. These additives are derived from plants containing phytogenic compounds as growth promoters, productivity, and livestock endurance enhancers [1].

Essential oils (EOs) are secondary metabolites of plants that have potential as feed additives because they are lipophilic, volatile, and possess antimicrobial properties [2]. The use of essential oils as feed additives in ruminant feed can modify rumen fermentation products and control harmful microbes [3]. These metabolites have been used as rumen modifiers due to their antimicrobial properties that inhibit the nutrient transport process by interfering with the cell membrane permeability to hamper microbial growth. This process reduces competition for nutrients as well as resources and allows beneficial microbes to thrive, leading to increased fermentation and volatile fatty acid (VFA) production in the rumen [4,5]. Recent meta-analysis results on the use of EOs as feed additives in vivo have shown improvement in rumen product fermentability, performance, and economic benefits in raising beef cattle [6]. Furthermore, Mclntosh et al. [7] found that the use of plant oil inhibited hyperammonia-producing bacteria.

Similar effects of EOs were found in fermentation and microbial activity during ensiling [8], where they demonstrated the ability to inhibit pathogenic bacteria and reduced deamination activity in silage during ensiling [9]. Furthermore, the effectiveness of EOs as additives during ensiling had not shown consistent results on nutrient quality and rumen fermentability. Several in vitro studies reported that the addition of EOs as additives during ensiling with certain types and levels can modify microbial fermentation, which has a positive effect on rumen fermentation [10]. However, some investigations revealed that EO supplementation reduced total VFA, dry matter digestibility, and organic matter digestibility [11]. There are also reports that EO supplementation did not affect rumen fermentation [12] but can increase the nutritional content of silage [13]. To generate a global conclusion, the differences in these results need to be further analyzed using statistical meta-analysis techniques by integrating data from each related study. Therefore, this study was conducted to evaluate the effects of EOs as additives on nutritional quality, fermentative products, and the microbial population of silage by integrating data from various related studies and analyzing them through meta-analysis. The in vitro rumen fermentation parameters were also integrated and analyzed.

2. Materials and Methods

2.1. Database Development

This study started with a literature search related to the use of EOs as additives for silage-making using the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) protocol [14]. Study papers were obtained through the Scopus database (https://www.scopus.com/search/, accessed on 27 February 2023) using the institutional IPB University access. The keywords used in the literature search were “silage” and “essential oils”, and a total of 107 articles were retrieved. Study papers were selected according to the following criteria: (1) articles published in English, (2) the presence of control treatment in the experiment without EOs, (3) the presence of EOs sources as additives silage during ensiling, and (4) the experiment on the rumen fermentability parameter was evaluated through in vitro rumen fermentation system. Based on the literature selection process, 17 articles that comprised 113 studies (comparisons) were included in this study. The schematic literature search and selection was shown in Figure 1, while the details of the studies included in the meta-analysis were summarized in

Table 1. Studies included in the database for meta-analysis regarding the effect of essential oil supplementation.

No	Reference	Year	Silage Ingredient	EOs Source	EOs Level (mg/kg)
1	Kung Jr. et al. [8]	2008	Corn	Mix (thymol, eugenol, vanillin, and limonene)	40–80
2	Foskolos et al. [9]	2015	Ryegrass	Eugenol, Cinnamaldehyde, Thymol, Carvacrol	50–2000
3	Chaves et al. [12]	2011	Barley	Cinnamon, Oregano, Sweet orange	37.5–120
4	Besharati et al. [13]	2021a	Alfalfa and apple pomace	Mix (ricinoleic acid, cardol, cardanol)	500
5	Termizkan et al. [15]	2011	Maize	Oregano	500–2000
6	Besharati et al. [16]	2021b	Alfalfa	Lemon seed	60–120
7	Besharati et al. [17]	2020	Leucerne	Cinnamon, Flaxseed, Lemon seed, Mix (cinnamon, flaxseed, lemon seed)	60–180
8	Salman et al. [18]	2018	Alfalfa	Cumin, Eugenol, Thymol	100–300
9	Hodjatpanah-montazeri et al. [19]	2016	Corn	Cinnamon, Thymol, Mint, Oregano, Cumin	120–240
10	Turan and Onenc [20]	2018	Alfalfa	Cumin	300–500
11	Akinci and Onenc [21]	2021	Vetch-oat	Cumin	200–500
12	Junior et al. [22]	2019	Sugarcane	Lemongrass	1000–3000
13	Duru [23]	2020	Alfalfa	Lavender flower	5–20
14	Soycan-Onec et al. [24]	2015	Field pea	Origanum, Cinnamon, Mix (Origanum and Cinnamon)	400
15	Li et al. [25]	2022	Mulberry	Amomum villosum	5
16	Besharati et al. [26]	2022	Alfalfa and apple pulp	Mix (ricinoleic acid, cardol, cardanol)	500
17	Vaičiulienė et al. [27]	2022	Perennial ryegrass, Red clover, and Blue alfalfa	Oregano and Thymol	20

.

The level, type, or source of EOs and the silage ingredients were compiled in a database, and the details are summarized in Table 1. Several articles provided multiple measurement units for particular parameters and data were converted into a pre-determined unit. The parameters integrated into the database were nutritional quality, namely dry matter (DM), pH, crude protein (CP), ether extract (EE), crude fiber (CF), crude ash (CA), neutral detergent fiber (NDF), acid detergent fiber (ADF), and acid detergent lignin (ADL), as well as fermentative products, including water-soluble carbohydrates (WSC), lactic acid, ethanol, acetic acid, propionic acid, butyric acid, succinic acid, NH3-N, weight loss (WL), total nitrogen, non-protein nitrogen (NPN). The microbial population was lactic acid bacteria (LAB), clostridia, yeast, and mold population. Meanwhile, the in vitro rumen fermentability parameters included rumen pH, ammonia (NH₃), methane emission (CH₄), 24 h gas production, total VFA, acetate (C₂), propionate (C₃), butyrate (C₄), iso-butyrate (isoC₄), valerate (C₅), iso-valerate (isoC₅), caproate (C₆), branched-chain volatile fatty acids (BCVFA), ratio acetate to propionate (C₂:C₃), dry matter digestibility (DMD), and organic matter digestibility (OMD).

2.2. Data Analysis

The data obtained were analyzed using the random effects meta-analysis method. The statistic of effect size calculation was based on the Hedges’d [28,29] using the formula below:

$$d=\frac{({\stackrel{\_}{X}}^E-{\stackrel{\_}{X}}^C)}{S}J$$

where ${\stackrel{\_}{X}}^E$ is the mean of the experimental group, ${\stackrel{\_}{X}}^C$ is the mean of the control group, and S is the pooled standard deviation, expressed as

$$S=\sqrt{\frac{(N^E-1)(s^E{)}^2+(N^C-1\left)\right(s^C{)}^2}{(N^E+N^C-2)}}$$

and J is the correction factor for the small sample size, expressed as

$$J=1-\frac{3}{(4\left(N^C+N^E-2\right)-1)}$$

where N^E is the sample size of the experimental group, N^C is the sample size of the control group, s^E is the standard deviation of the experimental group, and s^C is the standard deviation of the control group. A one-way random effect mathematical model is used in data analysis with the formula:

yi = θ + vi + εi

where the value of the effect size (in Hedge’s d) the i-th observation is (yi), the general parameter of the combined effect size is (θ), the actual effect size variation is (vi), and the error of the i-th observation is (εi). The estimated variation between studies (τ²) was measured using the DerSimonian and Laird method [30] with the following formula:

$${\tau }^2=\frac{Q-df}{C}$$

where Q is the weighted sum square, degrees of freedom is df, and the value of C (C). The meta-analysis software used in this study, namely the OpenMEE software, was derived from the work of Byron et al. [31] (http://www.cebm.brown.edu/openmee/, accessed on 10 January 2023). A cumulative forest plot (95% confidence interval) of the tested parameters was constructed using MedCalc software version 20.0.1.0 (https://www.medcalc.org/, accessed on 10 January 2023). Furthermore, the funnel plot and Egger’s test, which were employed to detect publication bias visually and statistically, were carried out using OpenMEE and MedCalc software.

3. Results

3.1. Nutritional Quality

Statistical descriptions of various parameters related to nutritional quality, fermentative products, microbial population, and in vitro rumen fermentation are presented in

Table 2. Descriptive statistics of the effect of EO supplementation on nutrient quality, fermentative product, microbial population, and in vitro rumen fermentability.

Variable	Unit	NC	Mean		SD		Min		Max
			Control	Treatment	Control	Treatment	Control	Treatment	Control	Treatment
Nutrient quality
Dry matter	%	52	29.0	29.7	1.09	1.08	22.2	22.6	34.9	35.1
pH		52	4.53	4.51	0.193	0.188	3.75	3.61	5.81	6.57
Crude protein	% DM	40	12.6	13.7	0.564	0.579	7.47	7.17	23.1	26.1
Ether extract	% DM	28	2.74	3.08	0.416	0.424	1.72	1.97	4.49	7.46
Crude fiber	% DM	8	30.4	29.2	0.137	0.088	21.0	18.7	41.5	39.9
Crude ash	% DM	16	10.5	9.97	0.484	0.311	3.88	5.37	15.6	13.5
NDF	% DM	40	50.6	49.3	1.05	1.03	33.2	28.2	64.3	64.9
ADF	% DM	30	31.3	30.3	1.06	1.05	21.0	15.3	53.3	47.5
ADL	% DM	11	7.84	7.55	0.311	0.319	6.66	6.29	9.51	12.7
Fermentative product
WSC	g/kg DM	23	6.68	6.25	0.328	0.353	3.97	3.12	14.1	15.3
Lactic acid	g/kg DM	33	50.7	51.6	6.86	6.81	23.3	3.10	69.8	83.1
Ethanol	g/kg DM	12	28.8	9.12	3.30	3.30	4.70	4.80	101	33.3
Acetic acid	g/kg DM	12	16.2	14.2	1.77	1.77	14.4	7.26	16.8	18.3
Propionic acid	g/kg DM	12	0.915	0.927	0.066	0.066	0.058	0.097	1.20	1.20
Butyric acid	g/kg DM	12	1.73	1.77	0.106	0.106	0.027	0.001	2.30	2.50
Succinic acid	g/kg DM	9	5.90	4.22	0.50	0.50	5.90	2.90	5.90	5.40
NH3-N	g/kg N	50	79.6	74.5	3.71	3.69	28.0	14.5	146	206
Weight loss	%	8	1.36	1.30	0.093	0.046	1.14	0.940	1.52	1.82
Total N	g/kg DM	12	42.1	41.0	2.35	2.35	38.6	37.0	48.9	49.8
NPN	g/kg DM	12	22.7	22.69	2.95	2.95	21.8	18.4	23.4	26.2
Microbial population
LAB	log10 CFU/g	29	6.28	6.70	0.507	0.513	0.00	2.17	8.58	8.53
Clostridia	log10 CFU/g	12	0.922	0.939	0.340	0.340	0.850	0.70	1.05	1.45
Yeast	log10 CFU/g	29	2.24	1.66	0.525	0.510	0.330	0.00	4.86	4.60
Mold	log10 CFU/g	29	1.49	1.01	0.555	0.538	0.300	0.00	4.87	5.05
In vitro rumen fermentability
pH		23	6.29	6.39	0.278	0.244	5.90	6.06	6.78	6.83
Gas 24 h	mL/g DM	32	172	167	27.4	34.2	103	88.7	194	220
CH4	mL/g DM	19	16.2	16.3	5.29	4.36	11.1	10.8	21.7	23.2
NH3	mmol/L	15	17.1	16.8	1.85	2.63	14.2	9.47	18.6	19.9
Total VFA	mmol/L	20	101	87.7	21.9	36.7	61.0	30.0	122	129
C2	mmol/L	14	59.5	62.0	5.64	2.56	49.5	58.4	67.3	68.1
C3	mmol/L	14	22.8	21.4	1.24	2.87	22.0	14.6	25.10	27.1
C4	mmol/L	14	12.7	11.8	4.62	3.14	5.52	4.95	20.6	16.3
IsoC4	mmol/L	5	1.17	1.05	0.216	0.357	1.00	0.70	1.52	1.61
C5	mmol/L	14	1.92	1.73	0.267	0.432	1.27	1.00	2.10	2.20
IsoC5	mmol/L	5	1.52	1.18	0.103	0.164	1.39	0.990	1.60	1.40
C6	mmol/L	9	0.420	0.450	0.00	0.037	0.420	0.390	0.420	0.510
BCVFA	mmol/L	9	2.90	2.93	0.00	0.169	2.90	2.70	2.90	3.30
C2:C3		9	2.80	2.74	0.00	0.069	2.80	2.60	2.80	2.80
DMD	%	22	58.3	55.2	16.9	17.6	24.0	23.0	74.4	79.0
OMD	%	16	55.0	49.5	16.2	15.4	27.2	26.1	68.6	62.6

NC: number of comparisons, SD: standard deviation, Min: minimum, Max: maximum, NDF: neutral detergent fiber, ADF: acid detergent fiber, ADL: acid detergent lignin, WSC: water-soluble carbohydrate, NPN: non-protein nitrogen, LAB: lactic acid bacteria, CH4: methane emission, VFA: volatile fatty acids, NH₃: ammonia, CH₄: methane, C2: acetate, C3: propionate, C4: butyrate, Iso-C4: iso-butyrate, C5: valerate, Iso-C5: iso-valerate, C6: caproate, BCVFA: branched-chain volatile fatty acids, C2:C3: ratio acetate to propionate, DMD: dry matter digestibility, OMD: organic matter digestibility.

. The results of the meta-analysis in

Table 3. Meta-analysis on the effect of essential oil supplementation on nutrient quality.

Variable	Unit	NC	Estimate	Lower Bound	Upper Bound	Std.Eror	p-Value	τ2	Q	Het.p-Value	I2
Dry matter	%	52	0.638	0.102	1.17	0.273	0.020	2.04	151	<0.001	74.2
pH		52	0.01	−0.472	0.49	0.246	0.967	2.15	208	<0.001	75.5
Crude protein	%DM	40	2.21	1.54	2.88	0.343	<0.001	3.00	169	<0.001	77.0
Ether extract	%DM	28	0.695	0.075	1.32	0.316	0.028	1.86	96.9	<0.001	72.1
Crude fiber	%DM	8	−3.63	−6.86	−0.400	1.65	0.028	13.8	55.3	<0.001	87.3
Crude ash	%DM	16	−0.452	−1.36	0.455	0.463	0.329	2.16	50.2	<0.001	70.1
NDF	%DM	40	0.087	−0.916	1.09	0.512	0.864	8.06	321	<0.001	87.9
ADF	%DM	30	−0.196	−1.05	0.656	0.435	0.652	3.97	165	<0.001	82.5
ADL	%DM	11	−1.02	−2.29	0.254	0.650	0.117	2.97	48.4	<0.001	79.3

NC: number of comparison, τ²: estimate of variance between studies in a random-effects meta-analysis, Q: study homogeneity, I²: percentage of variation across studies due to heterogeneity, NDF: neutral detergent fiber, ADF: acid detergent fiber, ADL: acid detergent lignin.

showed the comparison between controls and silage supplemented with EOs. Based on the nutritional quality, the addition of EOs increased (p < 0.05) several parameters of nutrient contents such as DM, CP, and EE, while crude fiber content decreased (p < 0.05) compared to that of the control. However, EO supplementation during ensiling did not affect the pH of the silage produced.

3.2. Fermentative Product and Microbial Population

The effects of EO supplementation on silage fermentative products and microbial population were compiled in

Table 4. Meta-analysis of the effects of EO supplementation on fermentative product and microbial population of silage.

Variable	Unit	NC	Estimate	Lower Bound	Upper Bound	Std.Eror	p-Value	τ2	Q	Het.p-Value	I2
WSC	g/kg DM	23	0.105	−1.02	1.23	0.574	0.855	5.22	130	<0.001	83.1
Lactic acid	g/kg DM	33	0.360	−0.334	1.05	0.354	0.309	2.83	160	<0.001	80.1
Ethanol	g/kg DM	12	−0.823	−2.29	0.639	0.746	0.270	5.80	101	<0.001	89.2
Acetic acid	g/kg DM	12	−0.983	−1.83	−0.135	0.432	0.023	1.69	45.0	<0.001	75.6
Propionic acid	g/kg DM	12	0.636	−0.022	1.29	0.336	0.058	0.868	30.7	0.001	64.2
Butyric acid	g/kg DM	12	−0.885	−2.06	0.289	0.599	0.139	3.59	76.8	<0.001	85.7
Succinic acid	g/kg DM	9	−2.54	−3.39	−1.69	0.435	<0.001	0.694	13.8	0.086	42.1
NH3-N	g/kg N	50	−1.86	−3.80	0.080	0.990	0.060	21.1	568	<0.001	91.4
Weight Loss	%	8	−1.09	−2.88	0.692	0.911	0.230	5.11	36.9	<0.001	81.0
LAB	log10 CFU/g	29	−0.308	−1.03	0.421	0.372	0.408	2.38	115	<0.001	75.7
Clostridia	log10 CFU/g	12	0.00	−0.407	0.407	0.208	0.999	0.00	6.39	0.846	0.00
Yeast	log10 CFU/g	29	−3.01	−4.67	−1.35	0.847	<0.001	11.0	221	<0.001	87.4
Mold	log10 CFU/g	29	−2.09	−3.42	−0.775	0.674	0.002	7.18	214	<0.001	86.9

NC: number of comparison, τ²: estimate of variance between studies in a random-effects meta-analysis, Q: study homogeneity, I²: percentage of variation across studies due to heterogeneity, WSC: water-soluble carbohydrate, NPN: non-protein nitrogen, LAB: lactic acid bacteria.

. Based on the analysis results, fermentative product variables such as acetic and succinic acid showed a significant decrease (p < 0.05) compared to the control. However, the production of WSC, lactic acid, and propionic acid showed insignificant results (p > 0.05). In microbial analysis, only mold populations had a significant effect (p < 0.05) on the resulting silage.

3.3. In Vitro Rumen Fermentability

The results of in vitro rumen fermentability in

Table 5. Meta-analysis on the effects of EO supplementation on in vitro rumen fermentation of silage.

Variable	Unit	NC	Estimate	Lower Bound	Upper Bound	Std.Eror	p-Value	τ2	Q	Het.p-Value	I2
pH		23	0.926	0.180	1.67	0.381	0.015	2.19	83.9	<0.001	73.8
Gas 24 h	mL/g DM	32	−0.992	−1.95	−0.037	0.487	0.042	6.03	277	<0.001	88.8
CH4	mL/g DM	19	1.02	−1.01	3.05	1.04	0.329	9.70	166	<0.001	89.1
NH3	mmol/L	15	0.596	−0.992	2.18	0.810	0.462	5.99	107	<0.001	87.0
Total VFA	mmol/L	20	−1.28	−2.36	−0.193	0.553	0.021	4.46	117	<0.001	83.8
C2	mmol/L	14	0.091	−0.634	0.815	0.370	0.806	1.17	36.5	<0.001	64.4
C3	mmol/L	14	−0.333	−1.01	0.348	0.347	0.337	0.951	32.5	0.002	60.0
C4	mmol/L	14	−0.161	−0.933	0.610	0.394	0.682	1.44	41.3	<0.001	68.5
IsoC4	mmol/L	5	−1.89	−3.75	−0.0170	0.953	0.048	3.34	17.5	0.002	77.1
C5	mmol/L	14	−0.423	−1.29	0.447	0.444	0.341	1.79	46.8	<0.001	72.2
IsoC5	mmol/L	5	−1.41	−2.26	−0.557	0.435	0.001	0.148	4.73	0.316	15.4
C6	mmol/L	9	0.973	−0.037	1.98	0.516	0.059	1.68	28.1	<0.001	71.5
BCVFA	mmol/L	9	0.121	−0.460	0.702	0.296	0.683	0.244	11.6	0.171	30.9
A:P		9	−0.430	−0.906	0.046	0.243	0.077	0.00	5.38	0.716	0.00
DMD	%	22	2.85	−1.20	6.89	2.07	0.168	48.7	292.9	<0.001	92.8
OMD	%	16	−3.27	−4.35	−2.20	0.547	<0.001	3.37	86.9	<0.001	82.7

NC: number of comparison, τ²: estimate of variance between studies in a random-effects meta-analysis, Q: study homogeneity, I²: percentage of variation across studies due to heterogeneity, VFA: volatile fatty acids, NH3: ammonia, CH₄: methane, C2: acetate, C3: propionate, C4: butyrate, Iso-C4: iso-butyrate, C5: valerate, Iso-C5: iso-valerate, C6: caproate, BCVFA: branched-chain volatile fatty acids, C2:C3: ratio acetate to propionate, DMD: dry matter digestibility, OMD: organic matter digestibility.

show that the administration of silage supplemented with EOs significantly affected several parameters. This led to a significant decrease in gas production, total VFA, isoC4 and isoC5, and organic matter digestibility (OMD) (p < 0.05) but did not affect methane production (p > 0.05). Meanwhile, the pH parameters showed significant results in terms of increasing (p < 0.05) rumen pH. These results indicate that EO supplementation was able to maintain rumen pH stability.

3.4. Sub-Group and Meta-Regression Analysis

The results of the sub-group analysis in Figure 2 based on mold population, which was the main parameter in identifying silage quality, show that only the EOs derived from cumin and oregano contributed to (p < 0.05) reducing mold populations during preservation. Meanwhile, the meta-regression results in Figure 3 revealed that the total population of mold negatively correlated with an increased level of added EOs with the following equation: Y = −1.67 − 0.001x. These results indicate that, as the level of use of EOs increased, the microbial population of spoilage decreased.

3.5. Publication Bias

The evaluation results of publication bias using the funnel plot test of the mold population parameter are presented in Figure 4. The funnel plot showed a clear asymmetrical image, and the statistical analysis of publication bias using Egger’s test yielded significant results (p = 0.0003). This indicated the presence of publication bias in the meta-analysis due to the different sources of EOs and silage substrates used in each paper. Therefore, this study used the random effect method and conducted a sub-group analysis based on the origin of EOs to determine their effect, as illustrated in Figure 2.

4. Discussion

Secondary metabolite-based additives have been widely used as alternatives for antibiotics due to their perceived safety and absence of residual matter in the resulting product. Secondary metabolites that can be used as silage additives were found to be plant-based EOs. This was because they possessed good antimicrobial activities and can inhibit the growth of pathogenic microorganisms during silage production. Moreover, pH and DM were the two most important factors in producing good silage [32] due to the importance of preventing low DM levels in acidic pH. Based on the nutrient quality analysis, the addition of EOs successfully increased the silage DM level without altering the pH level. This aligns with a study by Hodjatpanah-montazeri et al. [19], where silage supplemented with EOs from cinnamon extract had a higher DM value. This increase in DM was associated with a decreased population of microorganisms due to EO supplementation, leading to a minimal nutrient loss in silage [33]. Other nutrient quality-related parameters also showed a significant increase in CP and EE content, as well as high CP and EE levels in silage with added peppermint and oregano [19]. Moreover, the addition of EOs in silage reduced its crude fiber (CF) content compared to the control. This aligns with the work of Chaves et al. [12], who reported that the addition of EOs from cinnamon extract reduced the level of ADF and NDF.

Fermentation quality-related parameters were also analyzed in this study, as presented in Table 3. The addition of EOs during ensiling facilitated a significantly decrease in acetic and succinic acid production. The antibacterial activity of EOs was responsible for inhibiting the growth of acid-producing bacteria during silage fermentation. These results were in line with previous studies, where EOs limited the activities of acetic acid-producing bacteria such as enterobacteria [34,35]. Furthermore, EO supplementation caused a significant decrease in NH₃N production of the silage produced. Decreased NH₃N production indicated the ability of EOs in preventing silage proteolysis during ensiling. The use of other secondary metabolites such as tannin was also reported to inhibit proteolysis, as indicated by decreased NH₃N production, soluble N proportions, NPN, and free AA-N [36]. Proteolysis in silage can be influenced by several factors, such as feed type, forage species, forage harvest age, and the environmental conditions where silage is stored [37]. The presence of pathogenic microbes including clostridia and mold was also reported to affect increasing proteolysis [38] and deamination [36]. Therefore, the microbial population in produced silage was one of the important parameters included in this study. The results showed a decrease in mold and yeast population, indicating that the addition of EOs effectively inhibited the growth of pathogenic bacteria during ensiling. The use of Origanum vulgare and Thymus vulgaris extract was also proven to improve the hygienic quality of silage and prevent the concentration of mycotoxins, maintaining healthy livestock [22]. Some of the plant secondary metabolites, namely polyphenol, thymol, and carvacol were known to have growth inhibitory activity for several species of yeast, which can be associated with aerobic spoilage bacteria [39,40,41].

The analysis of in vitro rumen fermentability showed that the addition of EOs had a significant effect on the increase in rumen pH, which was also reported in previous studies [13,16,19]. The increase in the rumen pH was due to the decreased activity of lactic acid-producing bacteria (Lactobacillus sp.), which can reduce silage pH [42]. In the fermentation process, lactic acid-producing bacteria produced lactate, which was used by lactate-user bacteria to generate VFA in the form of propionate [43]. The antibacterial activity of EOs did not only inhibit harmful bacteria in silage but also the growth of VFA-forming bacteria. This condition led to a significant decrease in the total VFA concentration but not the main components, such as acetate, propionate, and butyrate. An increase in pH indicated a low level of feed fermentation, which was associated with a decrease in total VFA concentration [44]. High feed fermentation activity also caused a lower pH because lactic acid bacteria grew well, leading to an increase in the production of VFA [45]. Meanwhile, low feed fermentation activity caused pH to increase because the growth of lactic acid-producing bacteria was inhibited, which caused low VFA production [46]. The antimicrobial activity of EOs can also interfere with microbial activity in the rumen, thereby reducing the favorability of the rumen product.

The decreases in total VFA levels and some rumen fermentation products were caused by the antimicrobial activity of EOs against starch-degrading amylolytic bacteria. According to Sahan et al. [44], EO supplementation significantly inhibited the fermentative activity of rumen microbes and disrupted the overall feed fermentation process. Furthermore, the levels of BCVFA, which included IsoC4 and IsoC5, showed a significant decrease with EO supplementation. This was consistent with previous studies, where EOs significantly reduced BCVFA concentrations [44,47] and were associated with a decrease in NH₃ levels. The reduced production of NH₃ indicated that feed protein entering metabolic digestion was not immediately degraded by protease enzymes produced by rumen microbes [48]. BCVFA was a by-product of amino acid deamination in the rumen, including iso-butyrate, iso-valerate, and valerate. BCVFA in the rumen served as the main source of carbon chains for microbial growth. Proteolytic microorganisms hydrolyze feed protein into peptides in the rumen in preparation for amino acid breakdown. The low level of deamination caused a decrease in the production of BCVFA as a by-product. The antimicrobial activity of EOs and their protection of feed protein also contributed to a reduction in the concentration of rumen-degradable protein (RDP) and ammonia, leading to a decrease in microbial protein synthesis. Ammonia and BCVFA have been used as a source of nitrogen and carbon skeleton for microbial protein synthesis, respectively [49]. The decrease in BCVFA caused the availability of the carbon skeleton for microbial protein synthesis to reduce. This indicated that EO supplementation had a negative effect on feed fermentation in the rumen.

The digestibility of feed also experienced a significant decrease in the percentage of OMD. These results were consistent with previous reports, wherein a decrease in OMD occurred with the addition of EOs as feed additives [44,50] due to the degradability of organic matter—associated with silage fermentability. A significant decrease in OMD indicated that only a small amount of organic nutrients can be digested. According to Suharti et al. [51], the higher the digestibility value of the organic matter in the diet, the more nutrients can be digested to meet the nutritional needs of livestock. The decrease in OMD, along with the addition of essential oil levels, was due to the inhibition of the activity of bacteria fermenting feed in the rumen. This indicated that the antibacterial activity in EOs can be detrimental to the fermentability and digestibility of silage in vitro rumen fermentability. On the other hand, the effect of using different feed ingredients (forage or feed crop) can also affect the quality and digestibility of the silage used because the fiber content and active metabolite compound of each ingredient is different. In addition, the age of the plant also affects the nutrient content of silage and their digestibility [52]. The use of EOs as dietary supplements in in vitro [53] and in vivo [54] studies effectively increased several populations of fiber-degrading bacteria such as Fibrobacter succinogenes, Ruminococcus albus, and R. flavefaciens, thereby increasing the digestibility of feed in sheep. However, the results of a meta-analysis [9] on the use of EOs on digestibility found that supplementation did not affect the digestibility of protein, DM, organic matter, and fiber.

EO supplementation did not affect rumen methanogen production. However, some scholars have stated that EOs inhibit the growth of methanogen bacteria [53] and reduce greenhouse gas emissions [55]. The lack of significant effects was due to the antimicrobial activity, which did not significantly inhibiting the growth of methanogens, and the high availability of H₂ substrate from the fermentation of fibrous feed by bacteria in the rumen. The decrease in the population of protozoa must correspond to low methanogens due to the reduction in the symbiosis of methanogens with protozoa [56]. The inability of the methanogens to decrease even though their symbiosis with protozoan cilia was disturbed, was due to the larger size of protozoan, providing a suitable attachment for methanogens.

The sub-group analysis results based on the mold population in Figure 2 demonstrated that only EOs from cumin and oregano sources significantly inhibited the mold population during ensiling. This result aligned with the literature, which indicated a decrease in microbial population with oregano oil supplementation [57]. Furthermore, secondary metabolites such as EOs from the types of thyme and oregano are known to be good alternatives for protecting food or feed from pathogenic bacteria and preventing mold growth [58]. This statement confirmed the potent antimicrobial properties of EOs, with a broad spectrum that inhibited the growth of bacteria [59]. The significant effects observed among different types of EOs were attributed to their distinct compositions. This statement was supported by the results of an analysis of publication bias, which showed data bias, as indicated by an asymmetrical funnel plot image [60] and quantitative analysis using Egger’s test. This biased result can be attributed to several factors, such as the type of EOs and the different silage substrates used. Therefore, the method of analysis used in this study was a random effect model, which elaborated the results of each study. Based on the type of EOs, cumin and oregano showed more effective results in preventing the growth of pathogenic microbes such as mold, which can reduce the nutrient quality of silage.

5. Conclusions

The evaluation of EOs as silage feed additives during ensiling revealed the positive impact of EO supplementation on nutritional quality, fermentative products, microbial analysis, and in vitro rumen fermentation. The results showed that EO supplementation improved the nutritional quality of silage by increasing DM, CP, and EE. Furthermore, it effectively inhibited pathogenic microorganisms during ensiling, specifically mold and yeast populations. The antimicrobial activity of EOs affected microbial activity, thereby reducing fermentative products such as acetic and succinic acids. EO supplementation caused a decrease in in vitro rumen fermentability such as gas production, total VFA, iso-butyrate, and iso-valerate. There was also a positive impact on maintaining stable rumen pH, protecting feed protein by inhibiting the rate of NH₃N degradation during the fermentation process to reduce nitrogenous waste and increase nitrogen retention. Based on the sub-group analysis, it was discovered that EOs sourced from cumin and oregano had a significant effect on reducing mold population. This indicated that EO supplementation during storage effectively maintained nutritional quality and inhibited pathogenic microbes in silage during ensiling. However, the type and level of EOs used should be considered to avoid adverse effects on favorable rumen fermentation products.