Impacts of Chemical and Microbial Additives on the Quality of Forage Sorghum Silage During the Fermentation Process

Abstract

Additives are intentionally added to silage to reduce the growth of undesirable micro-organisms and to control the course of fermentation. This study aimed to evaluate the effect of two additives, a commercial product based on organic acids (OA) and Lentilactobacillus buchneri (Lb), alone or in combination with OA. The experiment was conducted in a 4 × 3 factorial completely randomized design, with five replicates per treatment, four additives (control, no additive (Control); commercial inoculant based on L. buchneri (Lb); additive based on organic acids (OA); Lb combined with OA (Blend)), and three fermentation periods (15, 30, and 90 days). The filamentous fungi count was higher in the Control silage during all fermentation periods. Lb silage showed greater aerobic stability (144 h) during all fermentation periods. The fermentation pattern was also influenced by inoculation; Lactobacillus was the most prevalent genus in Blend silage, and Lactiplantibacillus, Lacticaseibacillus, and Secundilactobacillus were predominant in OA silage, followed by Lentilactobacillus, which was higher in Lb silage. The addition of Lb and the Blend silage were the most efficient strategies, promoting greater accumulation of acetic acid and inhibiting yeasts, and the additives contributed to a more stable environment over 90 days of storage.

1. Introduction

Additives are utilized in silage to impede or diminish the proliferation of undesirable microorganisms, thereby promoting adequate fermentation and ensuring aerobic stability in addition to preserving the nutritional composition of silage [1]. This is particularly advantageous for forages, which, when harvested at their optimal stage, contain high levels of water-soluble carbohydrates (>150 g WSCs/kg DM) and low levels of dry matter (<250 g DM/kg natural matter), as is the case with forage sorghum, Sorghum bicolor (L.) Moench. In sorghum silage, the rapid conversion of sugars into lactic acid by lactic acid bacteria leads to a sharp decline in pH, which helps preserve the ensiled mass. However, under certain conditions, residual fermentable substrates may still support the proliferation of yeasts, especially when silage is exposed to air. These yeasts are primarily responsible for alcoholic fermentation and the aerobic deterioration of silage [2].

Although several microbial and chemical additives have been developed to address these issues by improving both the fermentation process and aerobic stability after silo opening [2,3,4,5], their benefits are not always significant [2,6,7,8,9].

The application of these additives at low rates may prove insufficient in preventing undesirable secondary fermentations. Additives containing organic acids, such as propionic, benzoic, and sorbic, exhibit high antifungal activity, leading to their extensive adoption to enhance the aerobic stability of silage [10]. Organic acids have an antimicrobial effect when they diffuse undissociated through the cytoplasmic membrane, releasing hydrogen ions into the cytoplasm and decreasing the intracellular pH [11,12].

Previous research has shown that the efficiency of these compounds is proportional to the total application in the ensiled biomass. Therefore, the incorporation of acids reduces the yeast population, prevents ethanol production, reduces fermentation losses, and prolongs aerobic stability, especially in corn silages [13]. In parallel, commercial inoculants based on obligate heterofermentative lactic acid bacteria, such as Lentilactobacillus buchneri, have been proposed for silage preservation. These inoculants produce organic acids with antifungal properties, including acetic and propionic acids, and inhibit the development of filamentous fungi and yeasts [14,15]. Kung, Jr. et al. [13] indicated that inoculants composed of L. buchneri could decrease yeast activity and consequently prolong aerobic stability due to the production of acetic acid. Although chemical inoculants and additives are widely used, some studies have compared the effectiveness of these products alone and in combinations for silage preservation [1,16].

In this context, we hypothesized that the application of a chemical additive, which is based on organic acids, and a commercial inoculant, which is based on L. buchneri, either alone or in combination, improves the fermentation profile, aerobic stability, and nutritional quality of forage sorghum silages throughout the fermentation period. The objective of this study was to evaluate the effects of Fresh CUT™ Plus and the inoculant L. buchneri alone or in combination with the chemical additive on the fermentation and nutritional characteristics, aerobic stability, and microbial diversity of forage sorghum silages as a function of the fermentation period.

2. Materials and Methods

2.1. Experimental Site

The experiment was conducted at the Forage Laboratory of the Department of Animal Science of the Federal University of Paraíba in Areia, State of Paraíba, Brazil, located at a latitude of 6°58′12″ South and longitude 35°42′15″ West, ±619 m above sea level. The forage sorghum cultivar BRS Ponta Negra (Guansafra Sementes^®, Fortaleza, State of Ceará, Brazil) was planted in an area of 0.76 hectares in the municipality of Riachão, State of Paraíba, Brazil, located at latitude 06°32′25″ South and longitude 35°39′35″ West, with an average altitude of ±175 m and average annual rainfall of 432 mm. According to the Köppen classification, the climate of the region is As’ (hot and humid semi-arid with rainfall from fall to winter) [17].

The soil of the experimental area was characterized as sandy soil with a light texture [18] and was prepared through plowing and harrowing and fertilized to meet the fertility requirements for sorghum crops. Forage sorghum was planted using a manual seeder (Semeadora Krupp^®, São Leopoldo, State of Rio Grande do Sul, Brazil). The sorghum was sown at a depth of approximately 1 cm and spacing of 0.10 × 0.80 m, with ten seeds per linear meter. Pre-sowing fertilization was performed with the application of 100 kg P₂O₅/ha as the single superphosphate (Fertilizantes Heringer^®, São Paulo, State of São Paulo, Brazil). Topdressing was performed at the V4 stage, using 45 kg N/ha as urea (Fertilizantes Heringer^®, São Paulo, State of São Paulo, Brazil).

2.2. Experimental Design, Harvesting, and Ensiling

The experiment was carried out in a completely randomized design in a 4 × 3 factorial arrangement, totaling 12 treatments (four additives × three fermentation periods) with five replicates per treatment, resulting in 60 experimental units. The four additive treatments consisted of control, with no additive (Control); a commercial inoculant based on Lentilactobacillus buchneri (Lb; NCIMB 40788—Lallemand^®, Santa Helena de Goiás, State of Goiás, Brazil), applied at 3 × 10¹¹ colony forming units (CFU)/g diluted in 500 mL of distilled water; an additive based on organic acids (OA; Fresh CUT™ Plus—Kemin South America, Indaiatuba, State of São Paulo, Brazil), applied at 0.5 kg/t of natural matter; and the combination of L. buchneri with OA (Blend). The three fermentation periods were 15, 30, and 90 days after ensiling.

Forage sorghum (BRS Ponta Negra) was harvested manually in July 2023. The harvested material was chopped to an average particle size of approximately 10 mm using a stationary forage chopper (EN9-F3B, Nogueira^®, Itapira, State of São Paulo, Brazil). After homogenization, the chopped forage was evenly distributed among treatments, and 60 experimental silos were prepared using 4 L plastic buckets (630EE, GrouPack^®, Cabedelo, State of Paraíba, Brazil), each measuring 55 cm in diameter and 19 cm in height. Silos were manually compacted to a density of approximately 600 kg/m³ of natural matter, sealed with 45 mm adhesive tape (3M-5802, Scotch^®, Sumaré, State of São Paulo, Brazil) and stored at ambient temperature until the respective opening day. The initial chemical composition of the fresh sorghum forage is presented in

Table 1. Characterization of the chemical composition and microbial populations of forage sorghum BRS Ponta Negra treated with chemical or microbial additives and their combinations before ensiling.

Item	Additive a
	Control	Lb	OA	Blend
Dry matter (g/kg natural matter)	273.80	252.55	253.20	253.30
Organic matter (g/kg DM b)	953.30	958.82	949.62	948.50
Mineral matter (g/kg DM)	46.69	41.18	50.37	51.50
Crude protein (g/kg DM)	70.93	52.85	55.62	61.61
Neutral detergent fiber (g/kg DM)	569.27	586.84	578.56	529.03
pH	5.87	5.90	5.75	5.69
Water-soluble carbohydrates (g/kg DM)	122.88	139.37	142.41	132.85
Buffering capacity (Emg NaOH/100 g DM) c	0.135	0.161	0.167	0.162
Lactic acid bacteria (CFU/g) d	5.38	5.36	6.14	5.98
Yeasts (CFU/g)	5.57	5.59	5.50	5.64
Filamentous fungi (CFU/g)	5.25	5.19	5.09	5.03

^a Additive: Control = control (no additive); Lb = commercial inoculant based on Lentilactobacillus buchneri NCIMB 40788, Lallemand^®, Brazil; OA: chemical additive based on organic acids Fresh CUT™ Plus, Kemin Industries^®, Brazil; Blend = combination Lb + OA. ^b DM = dry matter. ^c Buffering capacity (Emg NaOH/100 g DM): buffering capacity expressed in mmol/valency of sodium hydroxide for each 100 g of dry matter. ^d CFU/g: colony forming unit/g fresh forage on the logarithmic scale of base 10.

.

2.3. Fermentation Profile and Dry Matter Losses

For the analysis of the fermentation profile and quantification of microbial populations and organic acids, the 25 g samples of the plant and silages at 15, 30, and 90 days of fermentation were homogenized in 225 mL of saline solution (0.85 g NaCl (Dinâmica^®, São Paulo, State of São Paulo, Brazil)/100 mL of distilled water) in an industrial blender (LAR.2, Metvisa^®, Brusque, State of Santa Catarina, Brazil) for 1 min. The aqueous extract obtained was filtered through sterile cotton gauze and divided into three aliquots.

The first aliquot was used to measure the pH using a digital potentiometer (K39-1420A, Kasvi^®, São José dos Pinhais, State of Paraná, Brazil). Then, the second 50 mL aliquot of the aqueous extract was acidified (H₂SO₄ 1:1 v/v, Neon^®, Suzano, State of São Paulo, Brazil) and stored at −20 °C. To quantify the ammonia nitrogen (NH₃-N) content in relation to the total N, according to the method described by Bolsen et al. [19], a colorimetric method was used, with ammonium sulfate ((NH₄)₂SO₄) as the standard for calibration. The standard used for the determination of water-soluble carbohydrates (WSCs) by the concentrated sulfuric acid method was glucose, following the methodology described by DuBois et al. [20]. DM losses were quantified by the difference in weight of the buckets (final weight minus initial weight). It was calculated according to Magalhães et al. [21] by Equation (1):

DM losses (g DM/kg NM) = 1000 − [(DMab/DMfe) × 1000]

(1)

where DM losses (g DM/kg NM): dry matter losses in grams per kilogram of natural matter; DMab: kg dry matter of silage at silo opening (amount of silage in kg × % DM); DMfe: kg dry matter of forage at silo closing (amount of forage in kg × % DM).

2.4. Chemical Composition

Plant and silage samples were partially dried in a forced-air oven (55 °C/72 h). The samples were then ground in a knife mill (Wiley^®, Arthur H. Thomas, Philadelphia, PA, USA) with a 1 mm sieve for chemical composition analysis.

The plant, ingredient, and silage samples were analyzed for the quantification of DM (method 934.01), organic matter (OM, method 972.43), mineral matter (MM, method 942.05), crude protein (CP, N × 6.25, method 984.13), and ether extract (EE, method 920.39) according to AOAC [22]. The neutral detergent-insoluble fiber (NDF) content of these samples was also determined using thermostable α-amylase without sodium sulfite [23].

2.5. Microbial Populations and Organic Acids

The second aliquot of the aqueous extract was used to quantify the microbial populations of lactic acid bacteria, yeast, and filamentous fungi. Serial dilutions were made from 10⁻¹ to 10⁻⁸. Plating was performed in sterile Petri dishes (90 × 15 mm, FirstLab^®, São José dos Pinhais, State of Paraná, Brazil) using the pour-plate method in culture medium [24]. De Man, Rogosa & Sharpe Agar (K25-1043, Kasvi^®, São José dos Pinhais, State of Paraná, Brazil) was used for the cultivation of lactic acid bacteria (37 °C/48 h) and Dicloran Rose Bengal Chloramphenicol Agar (Oxoid™, Hampshire, UK) for the cultivation of yeasts (25 °C/72 h) and filamentous fungi (25 °C/120 h). For the quantification of colony forming units (CFU), plates containing 25–250 colonies were considered [25].

The third aliquot of the aqueous extract was used for the quantification of organic acids (lactic, acetic, propionic, and butyric acids). First, 10 mL of the aqueous extract was transferred to screw-capped test tubes (AJ03, PlenaLab^®, Jundiaí, State of São Paulo, Brazil) and vortexed (Vortex Mixer VX-200, LabNet^®, Oakland, NJ, USA), 2.0 mL metaphosphoric acid (Neon^®, Suzano, State of São Paulo, Brazil) was added, and the resulting solution was centrifuged (Mikro-120, Hettich Zentrifugen^®, Tuttlingen, Germany) at 13,000× g for 15 min. After this procedure, the supernatant was collected and placed in Eppendorf centrifuge tubes (Cral-Plast^®, Cotia, State of São Paulo, Brazil). The analysis of organic acids was performed using gas chromatography (GC) (Shimadzu, Richmond Hill, ON, Canada, SPD-10AVP), following the methodology described by Siegfried et al. [26].

2.6. Aerobic Stability

To assess aerobic stability, 1 kg of silage was collected when the silo was opened, weighed, and returned to the respective bucket without being compacted, where it was exposed to the air in a temperature-controlled room (25 ± 2 °C) for six consecutive days (144 h). A data logger (AK285, New AKSO^®, São Leopoldo, State of Rio Grande do Sul, Brazil) was placed in the center of the mass of each silage to record the temperature every 10 min. Three other data loggers were distributed throughout the room to measure the ambient temperature. Aerobic stability was defined as the number of hours the silage remained stable before its temperature increased by two degrees Celsius above the ambient temperature [27].

2.7. Bacterial 16S rRNA Gene Diversity

Bacterial community analyses were performed on the silages after 90 days of fermentation. Five biological replicates of each additive were used. Each sample, 25 g of macerated silage, was homogenized in 225 mL of standard buffered phosphate saline solution, filtered through sterile gauze, and centrifuged at 6000× g for 4 min at 4 °C, with the supernatant discarded and washed again in standard buffered phosphate saline solution. The pellet was resuspended in 900 µL of standard buffered phosphate saline solution and stored at −20 °C.

DNA extraction was performed with a commercial kit (Power Soil DNA Isolation Kit, MoBio, Carlsbad, CA, USA) according to the manufacturer’s protocol, using a 0.25 g pellet obtained by diluting 25 g of silage in 225 mL of saline solution (0.85% NaCl). The V3 and V4 hypervariable regions of the 16S rRNA gene were amplified by polymerase chain reaction: initial denaturation at 95 °C for 3 min, followed by 25 cycles of amplification, consisting of denaturation at 95 °C for 30 s, annealing at 55 °C for 30 s, and extension at 72 °C for 30 s, using primers 341F (5′-CCTACGGGNGGCWGCAG-3′) and 785R (5′-GACTACHVGGGTATCTAATCC-3′), which amplify a fragment of approximately 444 bp, using the Illumina MiSeq V2 kit (2 × 250 cycles).

Bacterial community diversity was analyzed using the QIIME2 v. 2023.2 platform and R v. 4.1.3 software [28]. The denoising of reads and the generation of amplicon sequence variants (ASVs) were performed using DADA2 [29], which includes quality filtering, chimera removal, and error correction. The quality profiles of forward and reverse reads were visually inspected, and trimming/truncation parameters were adjusted accordingly to remove low-quality bases. The Genome Taxonomy Database v. 207.0 was used for taxonomic assignment. To minimize biases associated with differences in sequencing depth, data were rarefied to the minimum sequencing depth observed across samples before diversity analyses. The alpha diversity indices calculated were Shannon, Simpson, and Pielou; for beta diversity, weighted and unweighted UniFrac distances were used [30,31].

2.8. Statistical Analysis

Data were analyzed using the Shapiro–Wilk test to check for normality and homogeneity of variance. The microbiological evaluation data expressed in CFU were converted to base ten logarithmic units (CFU/log.10). The data were analyzed according to a 4 × 3 factorial completely randomized experiment with four additives and three fermentation periods and with twelve factorial treatments and five replicates, according to the following statistical model:

Yỉʝκ = µ + ADỉ + FPʝ + (AD * FP) ỉʝ + ℮ỉʝκ

where Yỉʝκ = response variable; µ = overall mean; ADỉ = effect of additive ỉ (fixed effect); FPʝ = effect of fermentation period ʝ (fixed effect); (AD * FP) ỉʝ = effect of the interaction between AD and FP ỉʝ (fixed effect); ℮ỉʝκ = random residual experimental error.

All data were subjected to analysis of variance when found significant for the effect of the additive or fermentation period or interaction between both, in which the breakdown of these effects was performed. The means were compared by the Tukey test at a significance level of 5%, as a critical level of probability for type I error, using the Sisvar Software [32].

The alpha diversity of microbial communities was assessed using the nonparametric Kruskal–Wallis test to compare all groups, while the Wilcoxon–Mann–Whitney rank-sum test was used for pairwise comparisons between groups. To analyze the beta diversity, the permutational multivariate analysis of variance was applied with pairwise comparisons and 999 permutations to test the significance of the dissimilarities between the distance matrices of the samples and their respective groups. Furthermore, linear discriminant analysis of effect size was employed to examine the differential abundance of specific genera between treatments [33].

3. Results

3.1. Chemical Composition

Table 2. Significance (p-value) of the tested effects and the standard error of the mean (SEM) of the variables analyzed in forage sorghum BRS Ponta Negra silages treated with chemical or microbial additives and their combinations and ensiled for 15, 30, and 90 days.

Item a	p-Value b			SEM c
	Period	Additive d	Period × Additive
Dry matter (g/kg natural matter)	<0.001	0.0012	<0.001	1.745
Organic matter (g/kg DM)	0.122	<0.001	0.041	0.460
Mineral matter	<0.001	<0.001	0.017	0.013
Crude protein	<0.001	<0.001	<0.001	0.622
Neutral detergent fiber	<0.001	<0.001	<0.001	0.163
pH	<0.001	0.046	<0.001	0.045
Lactic acid bacteria	<0.001	<0.001	<0.001	0.042
Yeasts	<0.001	0.085	0.010	0.496
Filamentous fungi	0.002	0.004	0.843	0.276
Ammonia nitrogen	0.594	0.754	0.822	1.470
Lactic acid (g/kg DM)	<0.001	<0.001	<0.001	0.771
Acetic acid (g/kg DM)	<0.001	<0.001	<0.001	0.659
Lactic acid/acetic acid ratio	<0.001	<0.001	<0.001	0.024
Propionic acid (g/kg DM)	<0.001	<0.001	<0.001	0.917
Butyric acid (g/kg DM)	0.007	<0.001	0.023	0.001
Fermentation losses	0.122	<0.001	0.041	3.707
Water-soluble carbohydrates	<0.001	0.3224	0.0205	1.117
Aerobic stability	<0.001	<0.001	<0.001	5.454
Maximum temperature	<0.001	<0.001	<0.001	0.163
Minimum temperature	<0.001	0.046	<0.001	0.045
Average temperature	<0.001	<0.001	<0.001	0.042
Time (h) to reach maximum temperature	<0.001	0.085	0.010	3.383

^a DM = dry matter; h = hours. ^b Probability of effect (p ≤ 0.05) of fermentation period, additive, and interaction between fermentation period with additive (Period × Additive). ^c SEM = standard error of the mean. ^d Additive, Control = control (no additive); Lb = commercial inoculant based on Lentilactobacillus buchneri NCIMB 40788, Lallemand^®, Brazil; OA = chemical additive based on organic acids Fresh CUT™ Plus, Kemin Industries^®, Brazil; Blend = combination of Lb + OA.

summarizes the p-values for the evaluated variables, enabling the interpretation of the effects of the applied treatments.

The silage added with Lb showed the highest DM content at 30 days of fermentation (254.24 g/kg DM) before exhibiting a decline at 90 days. The lowest value was recorded in the Control silage at 15 days. The OM contents were similar among the silages at 15 days, except for the silage added with Blend. At 30 days, the Control and Blend silages had the lowest OM contents (945.03 and 946.08 g/kg DM), and at 90 days, the OA silage presented the lowest OM content (939.39 g/kg DM). The remaining silages did not differ from each other (

Table 3. Analysis of the interaction between fermentation period and additive on chemical composition of forage sorghum BRS Ponta Negra. The silages were treated with chemical or microbial additives and their combinations and ensiled for 15, 30, and 90 days.

Fermentation Period	Additive a
	Control	Lb	OA	Blend
	Dry matter (g/kg natural matter)
15 d	225.67 Cb	254.24 Ab	237.27 BCa	242.67 ABa
30 d	243.44 Ca	254.24 Aa	247.24 BCa	248.15 BCa
90 d	238.56 ABCab	224.15 Cc	241.45 ABa	244.95 Aa
	Organic matter (g/kg DM) b
15 d	949.22 Aa	949.78 Aab	949.57 Aa	944.56 Ba
30 d	945.03 Cb	952.17 Aa	949.59 ABa	946.08 BCa
90 d	948.78 Aa	947.97 Ab	939.39 Bb	947.97 Aa
	Mineral matter (g/kg DM)
15 d	50.78 Bb	50.22 Bab	50.42 Bb	55.44 Aa
30 d	54.97 Aa	47.83 Cb	50.41 BCb	53.92 ABa
90 d	51.22 Bb	52.03 Ba	60.61 Aa	52.03 Ba
	Crude protein (g/kg DM)
15 d	57.06 Da	71.22 Aa	57.35 CDa	62,76 BCb
30 d	56.32 Ba	72.15 Aa	58.14 Ba	69.48 Aa
90 d	58.69 BCa	59.17 BCb	55.45 Ca	65.12 Aab
	Neutral detergent fiber (g/kg DM)
15 d	629.40 Ab	606.60 Ab	636.59 Aa	621.02 Aa
30 d	713.45 Aa	628.42 Bab	630.35 Ba	614.69 Ba
90 d	640.24 Ab	674.95 Aa	633.53 Aa	622.34 Aa

^a Additive: Control = no additive; Lb = commercial inoculant based on Lentilactobacillus buchneri NCIMB 40788, Lallemand^®, Brazil; OA = chemical additive based on organic acids Fresh CUT™ Plus, Kemin Industries^®, Brazil; Blend = combination of Lb + OA. ^b DM: dry matter. Means followed by different uppercase letters in the same row are significantly different for the effect of additive within each fermentation period, and means followed by different lowercase letters in the same column are significantly different for the effect of fermentation period within each additive, both by Tukey’s test (p ≤ 0.05).

). The MM content was higher in the silage added with Blend at 15 days in the Control and Blend silages at 30 days (54.97 and 53.92 g/kg DM) and in the silage added with OA at 90 days (60.61 g/kg DM). As for the CP content, it was higher in the Lb silage at 15 and 30 days (71.22 and 72.15 g/kg DM, respectively) and higher in the Blend silage at 30 days. At 90 days, the Blend silage had the lowest CP content (65.13 g/kg DM). The NDF contents were stable at 15 and 90 days, but at 30 days, the Control silage presented the highest content (713.45 g/kg DM), while the treatments did not differ from each other (Table 3).

3.2. Fermentation Characteristics

The pH of the silages decreased over the course of fermentation regardless of the type of additive used. However, the silages treated with Lb exhibited higher pH values in the initial periods of fermentation (15 and 30 days), with values of 3.80 and 3.76, respectively. The Control silages presented lower pH values after 90 days of fermentation (

Table 4. Analysis of the interaction between fermentation period and additive on fermentation characteristics, microbial populations, and dry matter losses in forage sorghum BRS Ponta Negra silages. The silages were treated with chemical or microbial additives, alone or in combination, and ensiled for 15, 30, and 90 days.

Fermentation Period	Additive a
	Control	Lb	OA	Blend
	pH
15 d	3.59 Ca	3.70 Aa	3.67 BCa	3.64 Ca
30 d	3.63 Ba	3.76 Aa	3.64 Ba	3.60 Bab
90 d	3.46 Bb	3.51 ABb	3.58 ABa	3.50 ABb
	Lactic acid bacteria (CFU/g) b
15 d	6.96 Aa	6.97 ABba	6.81 ABa	7.03 Aa
30 d	6.87 Ba	7.53 Aa	6.50 Cb	6.65 BCb
90 d	6.07 Bb	6.41Ac	6.18 ABc	6.10 Bc
	Yeasts (CFU/g)
15 d	3.40 Aa	2.78 Aa	2.76 Aa	2.50 Aa
30 d	2.07 Ab	1.53 Bb	1.63 Bb	2.32 Aab
90 d	1.50 Ab	1.03 Bb	1.10 Ab	1.38 Ab
	Dry matter losses (g/kg DM) c
15 d	168.74 Aa	91.86 Ca	125.93 Ba	92.68 BCa
30 d	134.45 Ab	94.20 ABa	96.96 Bb	93.52 Ba
90 d	132.54 Ab	91.86 Ba	97.03 Bb	94.89 Ba
	Water-soluble carbohydrates (g/kg DM)
15 d	36.23 Ab	30.28 Ab	29.32 Ab	34.11 Aa
30 d	28.80 Bab	36.61 ABab	39.42 Aa	39.32 Aa
90 d	38.26 Aa	39.36 Aa	37.05 Aab	42.11 Aa
	Filamentous fungi (CFU/g) b
15 d	4.05	2.20 a	1.70	0.92
30 d	2.63	1.77 a	1.96	1.37
90 d	2.21	0.55 b	0.00	0.00
	Ammonia nitrogen (g NH3-N/kg total N)
15 d	15.84	17.48	18.15	21.46
30 d	18.58	15.89	17.86	13.30
90 d	16.29	19.70	16.26	28.91

^a Additive: Control = no additive; Lb = commercial inoculant based on Lentilactobacillus buchneri NCIMB 40788, Lallemand^®, Brazil; OA = chemical additive based on organic acids Fresh CUT™ Plus, Kemin Industries^®, Brazil; Blend = combination of Lb + OA. ^b CFU/g = colony forming unit/g fresh forage on the logarithmic scale of base 10. ^c DM = dry matter. Means followed by different uppercase letters in the same row are significantly different for the effect of additive within each fermentation period, and means followed by different lowercase letters in the same column are significantly different for the effect of fermentation period within each additive, both by Tukey’s test (p ≤ 0.05).

). Although statistically significant differences were observed, it is important to note that, from a biological standpoint, the variation in pH values among treatments is negligible and unlikely to impact silage quality. Higher values of DM losses were found for Control silages in all fermentation periods, with a decline over the course of fermentation, followed by OA silage at 15 days (125.93 g/kg DM) (Table 4). Higher values of WSCs were observed for Lb and Blend silages at 90 days of fermentation, with values of 39.36 and 42.11 g/kg DM, respectively, followed by OA and Blend silages at 30 days of fermentation, with no significant difference (Table 4).

3.3. Microbial Populations

It is worth noting that prior to ensiling, the Control treatment exhibited a higher CP value compared to the treated groups, indicating that the treatments may have influenced the preservation or degradation of protein content during fermentation. It is also noteworthy that this population decreased during the fermentation period, with a lower value for Blend silage at 90 days of fermentation, with an average of 6.10 CFU/g (Table 4).

The yeast population exhibited no change in average values for all silages during the 15-day fermentation period. However, the yeast population decreased during the ensiling process, reaching a minimum at 30 days for Lb silage (1.03 CFU/g). For the 90-day period, the silages Control, OA, and Blend remained with the same values; however, the silage Lb showed a low yeast population (Table 4). The population of filamentous fungi was higher in the Control silage across all fermentation periods, with a subsequent reduction at 90 days. Notably, the count of filamentous fungi was below detection levels at 90 days of fermentation for all silages added with additives (Lb, OA, and Blend) (Table 4).

3.4. Organic Acids

At 15 days of fermentation, all silages exhibited equivalent concentrations of lactic acid. At 30 days, lactic acid concentrations were lowest in the Control silage and highest in the same silage at 90 days (average 72.80 g/kg DM), followed by the Blend silage (60.16 g/kg DM). A similar trend was verified for acetic acid at 15 days, while at 30 and 90 days, the Lb silage presented the highest acetic acid contents (38.96 g/kg DM and 49.69 g/kg DM, respectively). The lactic acid/acetic acid ratio was comparable among the additives at 15 days, except for OA, which presented a lower lactic acid/acetic acid ratio. The Blend silage maintained a higher ratio throughout the fermentation process, with a value above 2 g of lactic acid for each 1 g of acetic acid produced (

Table 5. Analysis of the interaction between fermentation period and additive on organic acid contents of BRS Ponta Negra forage sorghum silages treated with chemical or microbial additives and their combinations and ensiled for 15, 30, and 90 days.

Fermentation Period	Additive a
	Control	Lb	OA	Blend
	Lactic acid (g/kg DM) b
15 d	65.45 Ab	62.07 Aa	65.17 Aa	65.26 Aa
30 d	49.15 Bc	54.34 ABb	57.13 Ab	58.41 Ab
90 d	72.80 Aa	57.27 BCab	56.27 BCb	60.16 Aab
	Acetic acid (g/kg DM)
15 d	32.29 ABa	32.76 ABc	35.31 Aa	31.38 ABa
30 d	26.74 Bb	38.96 Ab	28.29 Bb	26.18 Bb
90 d	35.68 Ba	49.69 Aa	27.94 Cb	27.13 Cab
	Lactic acid: acetic acid ratio
15 d	2.02 ABa	1.91 ABa	1.85 Ba	2.08 Aa
30 d	1.83 Bb	1.41 Cb	2.04 ABa	2.24 Aa
90 d	2.04 Aa	1.22 Bb	2.03 Aa	2.22 Aa
	Propionic acid (g/kg DM)
15 d	14.99 BCa	7.09 Cb	25.76 Aa	16.13 Bb
30 d	11.81 Aa	8.42 Ab	15.17 Ab	9.72 Ab
90 d	14.10 Ca	26.42 Ba	25.39 Ba	30.59 ABa
	Butyric acid (g/kg DM)
15 d	0.020 Aa	0.010 Ba	0.020 Aa	0.020 Aa
30 d	0.017 Aa	0.010 Ba	0.017 Aa	0.017 Aab
90 d	0.017 Aa	0.010 Ba	0.010 Bb	0.012 ABb

^a Additive: Control = no additive; Lb = commercial inoculant based on Lentilactobacillus buchneri NCIMB 40788, Lallemand^®, Brazil; OA = chemical additive based on organic acids Fresh CUT™ Plus, Kemin Industries^®, Brazil; Blend = combination of Lb + OA. ^b DM = dry matter. Means followed by different uppercase letters in the same row are significantly different for the effect of additive within each fermentation period, and means followed by different lowercase letters in the same column are significantly different for the effect of fermentation period within each additive, both by Tukey’s test (p ≤ 0.05).

).

OA silage showed the highest concentration of propionic acid at 15 days of ensiling. At 30 days, there was no difference between the additives. However, at 90 days, Blend silage presented the highest average for propionic acid. For butyric acid, Control silage presented the highest concentration throughout the fermentation period, while in Lb silage, the values remained constant (Table 5).

3.5. Aerobic Stability

The Lb silages exhibited greater aerobic stability (144 h) across fermentation periods and similar values for maximum and average temperature. At 30 and 90 days of fermentation, the Control, OA, and Blend silages did not differ from each other in terms of aerobic stability and maximum and average temperatures. At 90 days, the Control, OA, and Blend silages recorded the highest maximum temperatures (36.46 °C, 34.26 °C, and 35.50 °C, respectively) and the highest average temperatures of 28.30 °C, 28.53 °C, and 28.23 °C, respectively (

Table 6. Analysis of the interaction between fermentation period and additive on aerobic stability of BRS Ponta Negra forage sorghum silages treated with chemical or microbial additives and their combinations and ensiled for 15, 30, and 90 days.

Fermentation Period	Additive a
	Control	Lb	OA	Blend
	Aerobic stability (h) b
15 d	97.15 Ab	144.00 Aa	120.73 Aa	97.13 Ab
30 d	144.00 Aa	144.00 Aa	144.00 Aa	144.00 Aa
90 d	33.90 Bc	144.00 Aa	39.97 Bb	37.63 Bc
	Maximum temperature (°C)
15 d	25.12 Ab	25.06 Aa	24.90 Ac	24.80 Ac
30 d	25.14 Ab	25.38 Aa	26.32 Ab	26.36 Ab
90 d	36.46 Aa	25.58 Ca	34.26 Ba	35.60 ABa
	Minimum temperature (°C)
15 d	22.70 Cc	22.70 Cc	23.18 ABc	22.86 BCc
30 d	23.84 Ab	23.72 ABb	23.88 Ab	23.60 ABb
90 d	24.50 Aa	24.28 Aa	24.42 Aa	24.56 Aa
	Average temperature (°C)
15 d	24.32 Ab	24.36 Ab	24.40 Ab	24.20 Ac
30 d	24.63 Ab	24.54 Aab	24.71 Ab	24.53 Ab
90 d	28.30 Aa	24.83 Ba	28.53 Aa	28.23 Aa
	Time to reach maximum temperature (h)
15 d	53.82 Aa	36.52 Ab	36.35 Ab	30.17 Ac
30 d	55.65 Aa	93.66 Aa	84.25 Aa	94.28 Aa
90 d	68.12 Ba	96.62 Aa	70.77 ABa	62.60 Bb

^a Additive: Control = no additive; Lb = commercial inoculant based on Lentilactobacillus buchneri NCIMB 40788, Lallemand^®, Brazil; OA = chemical additive based on organic acids Fresh CUT™ Plus, Kemin Industries^®, Brazil; Blend = combination of Lb + OA. ^b h = hours. Means followed by different uppercase letters in the same row are significantly different for the effect of additive within each fermentation period, and means followed by different lowercase letters in the same column are significantly different for the effect of fermentation period within each additive, both by Tukey’s test (p ≤ 0.05).

).

3.6. Bacterial Diversity in the Silage

A total of 1.758 amplicon sequence variants were recovered following the denoising step for the bacterial datasets, which was sufficient to cover the bacterial diversity in all groups at 90 days of fermentation.

The alpha diversity indices showed that the Control silage presented a more diverse bacterial community, indicating both greater richness and evenness (Figure 1). In contrast, the control silage exhibited higher bacterial diversity, while the Lb, OA, and Blend silages showed a reduction in alpha diversity indices across all metrics. This reduction indicates a greater dominance of specific bacterial groups in the treated silages, which is desirable in this context, as it suggests a more selective and stable microbial community. The data further imply that inoculation alters microbial diversity, leading to a reduction in richness and evenness of the silages evaluated (Figure 1).

The unweighted UniFrac analysis demonstrated that the additives influenced the phylogenetic composition of the bacterial communities, as evidenced by the clear separation between the groups (Figure 2). The weighted index revealed that the relative abundances of the bacterial genera within these communities were more similar between some groups, except for the Blend silage, which remained distinct from the other treatments. The silage with organic acids showed significant differences on the microbial structure, differing from other silages (Figure 2).

The taxonomic depth of the bacterial groups was characterized by the prevalence of three phyla: Bacillota, Pseudomonadota, and Bacteroidota. The silages that were added with OA and Blend presented dominance of Bacillota, while the Control and Lb silages showed a more diverse bacterial community among the samples (Figure 3).

The family Lactobacillaceae was primarily responsible for fermentation, dominating the metataxonomic diversity in silages added with Lb, OA, and Blend. In contrast, the Control silage exhibited an increase in other taxa, such as Enterobacteriaceae, Burkholderiaceae, Pseudomonadaceae, Bacillaceae_D, Planococcaceae, and Clostridiaceae (Figure 4).

The fermentation pattern was also influenced by inoculation, as shown in the relative abundance of bacterial genera (Figure 5). The genus Lactobacillus was the most prevalent in Blend silages, while the genera Lactiplantibacillus, Lacticaseibacillus, and Secundilactobacillus were predominant in OA silage, followed by the genus Lentilactobacillus, which was more predominant in Lb silage. Control silage showed a shift in metataxonomic diversity, driven by Pseudomonas_E and other taxa not specialized in lactic acid bacteria, such as the genus Priestia (Figure 5).

As illustrated in Figure 6, the differential abundance analysis revealed distinct taxonomic profiles among the silages. The Control silage was characterized by a predominance of taxa such as Proteobacteria and Bacteroidota, whereas the silages treated with Lb and OA promoted specific genera within the Lactobacillaceae family. The Blend treatment also enriched for several Lactobacillus-related taxa, although with a different composition compared to the other treated silages. This LEfSe-based analysis (LDA effect size) highlights how the inclusion of microbial inoculants, chemical additives, or their combination selectively influenced the enrichment of bacterial groups. It is important to note that the differential abundance was assessed between treatment groups and does not reflect overall diversity indices such as alpha or beta diversity.

4. Discussion

The hypothesis that chemical additives based on organic acids and microbial inoculants reduce DM losses, improve the chemical composition, and increase the aerobic stability of silages was confirmed in this study.

4.1. Chemical Composition

Although the observed DM contents were lower than the recommended levels set by Kung Jr. et al. [13] for sorghum silages (280 to 320 g DM/kg natural matter),the fermentation process was not impaired. The reduction in DM content during storage can be attributed to the cumulative increase in dry matter losses, a phenomenon evident in Control silages, which presented the lowest DM contents.

Nkosi et al. [34] reported that DM levels of 250 g DM/kg natural matter are sufficient for achieving efficient fermentation, supporting the activity of the bacterial community in sorghum silages. Additionally, the lower DM losses in silages treated with chemical or microbial additives, compared to untreated silages, are consistent with the findings of Gandra et al. [35]. Notably, the microbial activity of L. buchneri can reduce the conversion of CP into soluble nitrogen compounds, minimizing its degradation during early fermentation (15 and 30 days). This process helps preserve CP levels in the silages with additives, resulting in minimal variation in ammonia nitrogen levels.

Moisture and storage time in silages without additives are key factors influencing protein losses, as they promote the activity of enterobacteria and, in some cases, Clostridium spp. In this context, the additives tested demonstrated effectiveness in preserving CP levels, with no significant differences between them [36]. Our findings corroborate with results reported by Nascimento et al. [37], who evaluated the effect of L. buchneri and sodium benzoate on sorghum silages and found that the additives contributed to the preservation of CP levels in the silage.

The silages treated with additives presented lower NDF content, possibly due to reduced degradation of non-fiber carbohydrates (such as sugars, starch, and pectins) compared to the Control silage. For additives based on organic acids, such as the OA additive, their action can accelerate the hydrolysis of structural compounds, including hemicellulose and pectins. This process increases their solubility and promotes their degradation [36]. As a result, the mineral fraction may increase, leading to a reduction in the relative proportion of OM, which explains the higher MM contents found in OA and Blend silages.

4.2. Fermentation Characteristics

All silages evaluated had pH values below 4, which indicates adequate fermentation. However, these values were lower than the ideal range suggested by Kung Jr. et al. [13]. Sorghum silages are naturally high in WSCs content and have low buffering capacity, resulting in a rapid drop in pH in the early stages of fermentation [37,38]. Despite this, all additives reduced the pH to levels that inhibit the growth of undesirable microorganisms in addition to promoting lower dry matter losses compared to the Control silage.

There were marked differences in fermentation efficiency between the additives tested. The Lb additive presented residual WSCs concentrations similar to those observed for the organic acid-based additives after 30 days of fermentation. However, at 90 days, all additives were effective in reducing losses and controlling secondary fermentations. Notably, the combination of Lb with OA, resulting in the Blend, promoted higher residual concentrations of WSCs possibly due to a dilution effect since organic acids limit the growth of some species of lactic acid bacteria and undesirable microorganisms. These findings corroborate with previous studies indicating that organic acid-based additives reduce or slow down the fermentation of WSCs compared to heterofermentative microbial additives [39,40].

4.3. Microbial Populations

The analysis of microbial populations indicated that the addition of the Lb additive to silage resulted in a predominance of lactic acid bacteria throughout the storage period (30 and 90 days), which favored the preservation of this microbial group. The initial increase in the population of lactic acid bacteria results from intense acidification, which stimulates their growth, as observed in Blend silage after 15 days of fermentation. However, as the fermentation period progresses, the population of lactic acid bacteria tends to stabilize due to the acidity of the medium [41].

Silages with a low pH showed an inhibitory effect on yeast growth, particularly those treated with Lb and OA. Notably, this acidification did hinder the degradation of protein into ammonia nitrogen, which created a favorable environment for lactic acid production [42]. The inhibitory effect can be attributed to the increased concentrations of acetic acid resulting from Lb, which limits yeast development. Additionally, the OA additive, composed of propionic, benzoic, and sorbic acids, exhibited strong antifungal properties [10]. The antimicrobial activity of these acids is due to their ability to permeate the cell membrane in an undissociated form, releasing hydrogen ions into the cytoplasm. This results in a decline in intracellular pH, leading to increased energy expenditure to maintain cellular homeostasis, ultimately inhibiting or disrupting microbial growth [43].

All additives reduced the population of filamentous fungi after 90 days of fermentation, with the Blend additive showing the best performance in controlling molds. Since silage fermentation is susceptible to proteolysis and fungal damage, especially during exposure to air, the tested additives composed mainly of propionic, benzoic, and acetic acid salts appear to have contributed to mitigating these effects [44].

4.4. Organic Acids

Silages treated with Lb exhibited lower levels of lactic acid and higher levels of acetic acid. This is due to the obligate heterofermentative metabolism of L. buchneri, which can convert lactic acid to acetic acid, 1,2-propanediol, propionic acid, and, to a lesser extent, carbon dioxide. This process justifies the lower lactic acid to acetic acid ratio observed in this study. Notably, at 30 and 90 days, the Lac/Ac ratio remained relatively low in the Lb treatment, likely due to increased acetic acid production resulting from the heterofermentative metabolism of Lactobacillus buchneri, which tends to shift fermentation towards acetic acid over time rather than lactic acid.

The propionic acid content was higher in silage treated with OA due to the high concentration of this organic acid in the additive. However, elevated levels of propionic acid were also observed in the Lb and Blend treatments. This suggests active in situ production of propionic acid, likely by Lactobacillus buchneri, which is known to convert lactic acid into acetic acid and 1,2-propanediol, with subsequent production of propionic acid under certain conditions. This pathway contributes to the improved aerobic stability observed in these treatments [45]. Previous studies have shown that propionic acid inhibits microbial growth, especially fungi, and under acidic conditions, it slows the growth rate and metabolic activity of yeasts, thereby prolonging aerobic stability [46,47].

To achieve effective silage fermentation, it is proposed that butyric acid levels should be kept below 5 g/kg DM [13]. In the present study, butyric acid concentrations were relatively low and did not have significant biological effect. These results agree with the findings reported by Wang et al. [48], who evaluated the use of organic salts as chemical additives in silage making, and Gheller et al. [49], who investigated the effects of propionic acid during the ensiling process. Both studies observed low levels of butyric acid in the silage, indicating an efficient fermentation process characterized by a predominance of lactic acid bacteria.

4.5. Aerobic Stability

Previous studies have shown that sorghum silages typically exhibit low aerobic stability after opening the silo, resulting in significant losses and increased costs due to the need to discard spoiled silage, which often contains fungi or other signs of degradation, as observed in this study for the Control silage. This instability is primarily linked to the large populations of yeast and the high levels of residual WSCs when the silage is exposed to air [50].

In the present study, the silage treated with the Lb additive showed improved aerobic stability, lasting 144 h in the evaluated fermentation periods. This suggests that the application rate of the additive was effective in controlling spoilage microorganisms, mainly through the production of acetic acid. In addition, silages added with OA and Blend showed improvements in aerobic stability compared to the Control silage. The effectiveness of these additives in improving aerobic stability is largely determined by the doses applied, which play a crucial role in improving fermentation characteristics of the silage.

The results obtained for OA and Blend are consistent with the findings of Silva Neto et al. [51], who demonstrated that the additive Mycoflake™, which contains acetic acid and propionic acid, also improved aerobic stability in corn grain silages. The antimicrobial properties of organic acids inhibit the growth of yeasts and filamentous fungi, which helps preserve the nutrients in silage during exposure to air. According to Li et al. [52], this inhibition of yeast activity results in acceptable levels of ammonia nitrogen and butyric acid, thereby reducing silage deterioration when in contact with oxygen.

In this context, Ranjit and Kung, Jr. [16] reported that applying L. buchneri at a dose of 1 × 10⁶ CFU/g provided greater stability to corn silage compared to an additive based on benzoate and propionate. Nevertheless, when applied at a lower concentration of 1 × 10⁵ CFU/g, this effect was diminished. Ranjit and Kung, Jr. [16] also noted that doses below 2 × 10⁵ CFU/g were insufficient to improve the aerobic stability of corn silage. In the present study, the application rate of Lb was 3 × 10⁵ CFU/g, which yielded satisfactory results. However, the lower aerobic stability observed in the Blend silage may be attributed to the fact that even when isolated, the OA additive maintained aerobic stability at 15 and 30 days. Thus, over extended periods, the additive may lose its antimicrobial efficacy, possibly due to the high balance of residual WSCs in the silage, which favors microbial activity in the ensiled material.

4.6. Bacterial Community Diversity

Alpha diversity was found to be high in Control silages. The Simpson and Pielou indices were lower in silages treated with Lb, indicating a reduction in microbial diversity, which is essential for successful ensiling [13]. The resistance to microbial succession observed in the inoculated treatments highlights the efficiency of these additives in improving fermentation quality.

Although variations exist in the composition of epiphytic communities in sorghum silages, the bacterial community that develops during fermentation differs significantly from the initial bacterial community [53]. As expected, the relative abundance of lactic acid bacteria increased after ensiling, becoming the dominant group in the later stages of fermentation.

In Control silage, the dominant bacterial genera were similar to those identified by Parvin et al. [54], Li and Nishino [55,56], and Ni et al. [57]. These previous studies reported that Enterobacter, Pseudomonadaceae, and Clostridiaceae as the primary bacterial groups found in high-moisture silages. These microorganisms are well adapted to moist and low-acidity environments: Enterobacter and Pseudomonadaceae are facultative anaerobes that rapidly metabolize sugars under low oxygen tension, while Clostridiaceae, strict anaerobes, ferment amino acids and carbohydrates into undesirable compounds. This microbial composition reflects an undesirable fermentation process in high-moisture silages. The presence of Clostridium in silage is challenging, as its metabolic activities convert sugars and proteins into butyric acid, which negatively affects fermentation quality.

Changes in the composition of dominant bacterial taxa appear to be correlated with their physiological characteristics and their adaptation to environmental changes induced by additives. This correlation explains the greater dominance of Lactobacillus in silages treated with the combination of additives, as in the case of the Blend additive. In contrast, silages treated with the OA additive showed a prevalence of Lactiplantibacillus, Lacticaseibacillus, and Secundilactobacillus. The bacterial genera identified in our study demonstrated a positive correlation with the production of lactic, acetic, and propionic acids.

Moreover, the similar abundance of Lactobacillus, Lactiplantibacillus, and Lacticaseibacillus, combined with the higher fermentative quality of silages amended with Blend and Lb, can be attributed to the increased production of lactic acid by exogenous lactic acid bacteria compared to the epiphytic lactic acid bacteria present in the Control silage. This phenomenon accounts for the lower levels of lactic acid and the higher proportion of acetic acid in these silages, indicating a predominantly heterofermentative fermentation process.

Previous studies have shown that both organic acid-based additives and L. buchneri-based inoculants enhance the production of acetic and propionic acids. These acids inhibit the activity of undesirable microorganisms, including some lactic acid bacteria [56,58,59]. Nascimento et al. [37] reported that combining chemical and microbial additives can optimize the population dynamics of Lactobacillus. This combination promotes a significant increase in these bacteria at the beginning of fermentation and extends their activity throughout the storage period.

It is important to highlight that Lactobacillus, Lactiplantibacillus, and Lacticaseibacillus play significant roles in modulating the microbiota during fermentation through the production of antimicrobial compounds. These metabolites not only contribute to a reduction in pH but also display an antagonistic effect against various acid-intolerant bacteria. This enhances the aerobic stability of silage, optimizes the preservation of nutrients, and improves the overall quality of fermentation.

5. Conclusions

Microbial and chemical additives enhance the quality of forage sorghum silage, reduce dry matter losses, preserve chemical composition, and increase the aerobic stability of silage. The incorporation of L. buchneri and Fresh CUT™ Plus, either individually or in combination, represents a highly effective strategy for preparing sorghum silage. This approach results in a substantial increase in acetic acid accumulation while effectively inhibiting the proliferation of yeasts and filamentous fungi. Such modifications positively influence fermentation dynamics, contributing to a more stable silage environment throughout the 90 days of storage. In addition, the assessment of the microbial diversity profile demonstrates the beneficial effects of these additives on silage preservation, which confirms their significance in maintaining silage quality.