Cottonseed Cake as a Feed Supplement: Effects on Nutrient Intake, Digestibility, Performance, Nitrogen Balance, and Ruminal Profile of Lambs Fed Sugarcane Silage-Based Diets

Abstract

Using additives in sugarcane silage can reduce dry matter losses and enhance animal performance by preserving nutritional value. This study evaluated the performance, nutrient digestibility, and blood parameters of sheep fed sugarcane silage with or without cottonseed cake. Twenty-six uncastrated, mixed-breed male lambs (approximately 6 months of age; 26 ± 1.3 kg) were allocated to two dietary treatments based on sugarcane silage (SS) and cottonseed cake (CSC), which differed in the form of feed presentation: (1) the control, consisting of SS and fresh CSC provided simultaneously but offered separately, without physical mixing; and (2) the Partial Mixed Ration silage (S + CSC), in which SS and CSC were pre-mixed and ensiled together at a proportion of 80:20 (natural matter basis). Dry matter intake (1620 g/kg) was similar between diets, but dry matter digestibility (64.75%) and average daily gain (202.88 g/day) were higher in the cottonseed cake group, leading to greater total weight gain (8.11 kg). These animals also had a lower acetate/propionate ratio (4.2 vs. 2.0 mmol/L) and higher blood glucose (44 vs. 35 mg/dL). Higher N intake, urinary N, and retained N were observed in the cottonseed cake diet, which also improved the diet’s mineral balance. In conclusion, adding cottonseed cake to sugarcane silage enhances fermentation, preserves nutritional value, and improves sheep performance.

1. Introduction

Sugarcane is a competitive source of tropical forage due to its high dry matter (DM) yield, which can exceed 30 t DM/ha [1]. It is a high-yield forage crop rich in water-soluble carbohydrates (WSCs), such as sucrose, making it a promising energy source for ruminants [2,3]. However, its low crude protein content and limited fiber digestibility can compromise animal performance if not properly supplemented. Moreover, the high WSC content, while beneficial for fermentation, can also promote yeast growth and lead to dry matter (DM) losses during the ensiling process [3]. Carvalho et al. [4] stated that the characteristic fermentation pattern of sugarcane results in alcoholic fermentation, representing a considerable challenge for its use as forage. Faced with these limitations, alternatives have emerged, such as the use of additives, seeking to minimize adverse effects and improve the quality of sugarcane silage.

Different additives are known for their use in silage production, including to improve fermentation and prevent spoilage by aerobic bacteria, filamentous fungi, and undesirable yeasts [5,6]. Among the additives used, nutritional additives, in particular, stand out because they are a fermentative and nutritional combination, resulting in benefits when ensiled with concentrates compared to the use of inoculants [5]. An example of an additive that can be used is cottonseed cake, easily available on the market in many countries. Recently, Justino et al. [7] investigated the use of cottonseed cake as an additive in sorghum silage and found improvements in nutritional value, fermentation profile, DM recovery, aerobic stability, and a reduction in the yeast population in the material. Santana et al. [8] concluded that the addition of cottonseed cake to millet silage reduced losses and improved the fermentation profile and nutritional value of the silages.

Thus, the practice of ensiling concentrates with forages is a technology called Partial Mixed Ration (PMR). Fermentation in ensiled PMR increases the digestibility of concentrates, as observed in previous studies, with increased digestibility of proteins, starch, and fibers due to acid hydrolysis [9,10].

Furthermore, the additive effect of protein concentrates is hypothesized to preserve the energy value of sugarcane, controlling alcoholic fermentation, and thus maintaining its energy value. Therefore, animals fed diets containing PMR of sugarcane silage added with cottonseed cake present better performance compared to those fed exclusive sugarcane silage with concentrate supplied fresh. Therefore, this study aimed to evaluate the productive performance, nutrient intake, digestibility, nitrogen balance, and ruminal and blood parameters of sheep fed diets containing PMR with cottonseed cake compared to sugarcane silage with cottonseed cake supplied fresh.

2. Materials and Methods

2.1. Location, Experimental Facilities, and Period

This study was carried out on a private property located in São José dos Cordeiros, state of Paraíba, Brazil (7º39”07′ S, 36º80”84′ W, 545 m altitude above sea level). The climate of the region is classified as BSh (hot semi-arid), characterized by summer rainfall, with an average annual rainfall of 764.4 mm and an average annual temperature of 27.6 °C [11].

This study was performed following a research protocol approved by the Ethics Committee on Animal Experimentation of the Federal University of Paraíba (Protocol 8681210920/2020). The animals were confined for 54 days (14 days for adaptation and 40 days for data collection). The animals were housed in individual stalls equipped with feeders and drinkers, in a masonry shed with concrete floors. Water and mineral supplement were available ad libitum. The animals were weighed every 15 days.

2.2. Experimental Design

This was a completely randomized experimental design with two treatments and 13 repetitions for each treatment, resulting in a total of 26 experimental units. Treatment one (control) consisted of sugarcane silage (SS) and cottonseed cake (CSC) offered separately but simultaneously; the CSC was provided fresh, without prior fermentation or mixing with the silage.

Treatment two consisted of a Partial Mixed Ration silage (S + CSC), composed of 80% sugarcane and 20% cottonseed cake (on a natural matter basis), which were pre-mixed and ensiled together, resulting in a single silage offered to the animals. Both treatments maintained the same proportion of cottonseed cake in the diet; however, the key difference was the physical form and fermentation: in Treatment 1, the cottonseed cake remained unfermented (fresh), while in Treatment 2, it was ensiled together with sugarcane.

2.3. Animals, Diets, and Management

Twenty-six uncastrated, mixed-breed male lambs, approximately six months of age and with an average initial weight of 26 ± 1.3 kg, were used in the experiment. During the first fourteen days, the animals underwent a period of adaptation to the diet and were vaccinated against clostridial diseases (Biovet^® Resguard Multi, Vargem Grande Paulista, SP, Brazil) and broad-spectrum oral antiparasitic treatment (Diantel^®, Closantel sodium, Madrid, Spain) [12]. The animals were housed in individual pens measuring 1 × 1.5 m, where measurements of intake, refusals, feces, and urine were taken every seven days.

Ensiling was carried out on the farm, the harvest was performed manually using machetes, and the collected material was processed through stationary forage (PP-35, Pinheiro máquinas, Itapira, São Paulo, Brazil) from where sugarcane was mechanically harvested to approximately 2 cm. Cottonseed cake was then manually mixed with the chopped sugarcane according to the respective treatment. The mixture was ensiled in double-layer polyethylene bags (polyethylene on the inner side and polyester on the outer side), each with a capacity of approximately 90 kg. The material was manually compacted using wooden tampers to remove as much air as possible and ensure anaerobic conditions. After sealing, the silage bags were stored in a shaded and ventilated shed at ambient temperature and allowed to ferment undisturbed for 60 days.

Diets consisted of silages, ground corn, cottonseed cake, soybean meal, urea/ammonium sulfate (9:1) with 48% nitrogen and equivalent to 300% crude protein, mineral pre-mix (Phos Ovinos^®, Tortuga, São Paulo, Brazil) containing 115 g/kg calcium; 60 g/kg phosphorus; 100 mg/kg cobalt; 600 mg/kg copper; 30 g/kg sulfur; 1200 mg/kg iron; 600 mg/kg fluorine; 130 mg/kg iodine; 10 g/kg magnesium; 600 mg/kg manganese; 12 mg/kg selenium; 175 g/kg sodium; 4000 mg/kg zinc; and ammonium chloride. The diet included ground corn (dried corn kernels ground using a hammer mill fitted with a 3 mm screen) as an energy source. Analyses of the chemical composition of diets and ingredients were carried out (

Table 1. Chemical composition of ingredients used in formulation of experimental diets (g/kg⁻¹ DM).

Item	SS	S + CSC	GC	SM	CSC
Dry matter	245.00	415.80	900.20	912.80	945.00
Organic matter	919.00	950.40	984.30	909.00	960.70
Ash	19.80	21.90	14.10	82.20	37.10
Crude protein	30.00	122.70	98.00	424.0	228.10
Ether extract	14.30	55.50	50.70	22.00	88.70
NFC	269.50	219.00	705.50	295.80	207.50
WSC	288.50	167.30	-	-	79.34
NDFom	666.40	580.90	131.70	176.00	438.60
ADFom	345.40	323.10	25.00	82.10	324.30
HEM	321.00	257.80	106.70	93.90	114.30
ADL	78.40	101.30	3.80	7.90	71.20

SS: solely sugarcane (control); S + CSC: sugarcane + 20% cottonseed cake (as fed basis); GC: ground corn; SM: soybean meal; CSC: cottonseed cake. NFC: non-fiber carbohydrates; NDF: neutral detergent fiber corrected for residual ash and protein; WSC: water-soluble carbohydrates; ADF: acid detergent fiber corrected for residual ash and protein; HEM: hemicellulose; ADL: acid detergent lignin.

).

Lambs received PMR placed directly in the feeders twice daily, at 8:00 a.m. and 4:00 p.m. The roughage/concentrate ratio of the diets was kept at 50:50 to achieve an average daily weight gain of 200 g, calculated according to NRC [13] (

Table 2. Proportion of ingredients (%) and chemical composition of experimental diets (g/kg⁻¹ DM).

Item	Diets
	SS	S + CSC
Proportions
Silage	47.03	70.59
Ground corn	26.06	23.99
Soybean meal	2.72	3.16
Cottonseed cake	21.56	0.00
Urea/ammonium sulfate (9:1)	0.51	0.19
Phos Ovinos	1.15	1.13
Chloride ammonium	0.97	0.94
Chemical composition
Dry matter	560.90	604.60
Organic matter	920.60	935.80
Ash	23.30	21.40
Crude protein	135.80	145.40
Ether extract	39.70	52.00
Non-fiber carbohydrates	333.20	363.40
NDFom	447.10	447.20
ADFcp	241.10	236.70
Hemicellulose	206.00	210.60
Acid detergent lignin	72.70	53.40
Metabolizable energy (Mcal kg−1 DM)	3.63	3.85

SS: solely sugarcane (control); S + CSC: sugarcane + 20% cottonseed cake (as fed basis); NDF: neutral detergent fiber corrected for residual ash and protein; ADF: acid detergent fiber corrected for residual ash and protein.

). A 10% margin for leftovers was allowed, with daily adjustments made to ensure ad libitum intake.

2.4. Growth Performance and Nutrient Utilization

At the beginning and end of the experimental period, the animals were subjected to a 16 h solid fasting and were weighed on an electronic scale (Welmy, W 300, Santa Bárbara d’Oeste, PR, Brazil). The weight differences between the initial body weight (IBW) and the final body weight (FBW) were used to calculate the total weight gain (TWG), average daily weight gain (ADG), and feed efficiency (FE). Nutrient intake was estimated by comparing the average total nutrient content in the diet offered with the total nutrient content in the leftovers of each animal. The ADG was calculated based on the weight differences between the IBW and FBW of the animals by the number of experimental days. FE was determined by dividing the average daily gain (g/day⁻¹) by the DM intake (g/day⁻¹).

Digestibility coefficients (DCs) for DM, crude protein (CP), ether extract (EE), organic matter (OM), neutral detergent fiber (NDF), total carbohydrates (TCs), and non-fiber carbohydrates (NFCs) were calculated using the following equation: DC = [(kg fraction ingested − kg fraction excreted)/(kg fraction ingested)] × 100 [14]. The digestible energy (DE) was initially quantified as the product of the TDN content and the factor 4.409/100, considering an ME concentration of 82% of DE.

Nutrient digestibility was calculated based on the collection of feed, leftovers, and feces from animals in individual folds. Fecal DM output was estimated using indigestible neutral detergent fiber (iNDF) as an internal indicator following the methodology of Casali et al. [14]. To determine iNDF concentrations, 0.5 g samples of feed, feces, and leftovers were incubated for 288 h in the rumen of a fistulated bovine on a standard diet.

2.5. Sampling Procedures

Diet, leftovers, feces, and urine samples were taken weekly from animals housed in individual folds. Leftovers and feed samples were collected in the morning, before feeding. Feces were collected twice a day (spot collection), at different times each day (Day 1: 6 a.m. and 2 p.m.; Day 2: 8 a.m. and 4 p.m.; Day 3: 10 a.m. and 6 p.m.). Feces were collected using the spot method, and each sample was promptly identified and frozen. At the end of the period, all fecal samples from each animal were combined for later analysis.

Blood and urine samples were collected weekly from animals in individual folds, four hours after morning feeding. Blood samples were drawn into vacuum tubes (Vacutainer^®, BD Inc., Franklin Lakes, NJ, USA) and refrigerated until centrifugation (1200× g for 15 min) to obtain plasma, which was transferred to microtubes (1.5 mL) and frozen for glucose and urea analysis. Urine samples were filtered, identified, and acidified. Aliquots of 10 mL of urine were diluted in 40 mL of 0.036 N sulfuric acid.

Ruminal fluid was collected on the last day of confinement from six randomly selected animals of each treatment, using an esophageal probe connected to a suction pump, four hours after feeding. Due to the invasive nature of the collection, which could cause stress to the animals and potentially affect their intake and weight gain, it was not performed weekly. The collected material was filtered through gauze to extract the liquid fraction, and 2 mL of 25% metaphosphoric acid was added to each sample to stop any ongoing fermentation and prevent the volatilization of compounds. Samples were centrifuged at 17,000× g for 10 min, and the supernatant was pipetted and then frozen for later analysis.

2.6. Chemical Analysis and Calculations

Silage samples weighing approximately 300 g were analyzed for fermentation profile and chemical composition. Silage pH was measured in duplicate for each experimental repetition, according to Jobim et al. [15]. The determination of ammonia nitrogen (N-NH₃) in silages and ruminal fluid followed the methodology described by Chaney and Marbach [16]. The water-soluble carbohydrate (WSC) content—comprising primarily simple sugars such as glucose, fructose, and sucrose, which are readily fermentable by lactic acid bacteria—was determined as proposed by Dubois et al. [17] with the extraction method described by Corsato et al. [18]. For the analysis of ethanol and organic acids in silages and ruminal fluid, the samples were processed according to the method described by Siegfried et al. [19] and analyzed by gas chromatography using different temperature ramps.

Ruminal fluid samples were analyzed by high-performance gas chromatography (Ciola Gregory^®, model CG Master). The initial oven temperature was programmed to start at 50 °C, held for 1 min, followed by heating at 10 °C min⁻¹ to 110 °C. Silage samples were analyzed with the initial oven temperature programmed to start at 60 °C, held for 1 min, followed by heating at 14 °C min⁻¹ to 200 °C. This temperature was held for 4 min. In both cases, a flow rate of 2 µL was used in a 1:20 split mode, with a J&W Carbowax column (30 m; 0.25 mm × 25 µm). The injector temperature was 220 °C and the detector temperature was 230 °C. Hydrogen was used as carrier gas at a flow rate of 1 mL min⁻¹. Organic acid standards were obtained using Sigma-Aldrich commercial kits.

Silages, concentrates, total diets, leftovers, and feces were dried in a forced-air oven at 55 °C for 72 h and ground to 1 mm using a Wiley Mill (Marconi, Piracicaba, Brazil). The analyses were conducted at the Animal Nutrition Laboratory of Embrapa Semiárido, located at BR-428, Km 152, Zona Rural, Petrolina, Pernambuco, Brazil. Determinations were made according to the methodologies proposed by Detmann et al. [20]: dry matter (DM; method G-003/1), ash (method M-001/1), crude protein (CP; method N-001/1), and ether extract (EE; ANKOM XT10 device, according to the manufacturer’s manual). Crude protein was calculated by multiplying total nitrogen by 6.25.

Neutral detergent fiber corrected for ash and residual protein (NDFom; method F-002/1) was determined according to Mertens et al. [21]. Acid detergent fiber (ADF; method F-004/1) was determined according to Van Soest et al. [22]. Acid detergent lignin (ADL) was estimated by the method of Robertson and Van Soest [23]. To estimate total carbohydrates (TCs), we used the equation proposed by Sniffen et al. [24]: TC (g/kg⁻¹ DM) = 1000 − (CP + EE + Ash). To calculate non-fiber carbohydrates (NFCs), the equation recommended by Hall [25] for foods containing urea was adopted due to its presence in the diet offered, NFC (g kg⁻¹ DM) = 1000 − [CP − (CP urea + Urea)) + NDFom + EE + Ash], where CP urea and NDFom represent crude protein from urea and neutral detergent fiber corrected for ash and protein, respectively.

Blood samples were analyzed for glucose and urea concentrations in a specialized laboratory. The pH of ruminal fluid samples was analyzed immediately after collection using a portable digital potentiometer (KASVI^®, model K39-0014, São José dos Pinhais, PR, Brazil). Ruminal volatile fatty acids (VFAs) were analyzed by high-performance gas chromatography as previously described. Nitrogen balance (NB) was determined by the following formula:

NB = (supplied N − leftovers N) − (fecal N + urine N).

(1)

2.7. Statistical Analysis

The data were analyzed using the SAS^® statistical package version 9.0 (SAS Inc., Cary, NC, USA) (SAS, 2004) [26]. The data related to the silage fermentation profile were tested by analysis of variance using the GLM procedure according to the following model:

Yi = μ + αi + ei

(2)

where

Yi = observed value; μ = overall mean; αi = fixed effect of cottonseed cake inclusion level; ei = residual error.

Data on nutrient intake and digestibility, performance, nitrogen balance, and ruminal and blood parameters of the animals were analyzed using the MIXED procedure considering the diets as fixed effects and the animals as random effects, according to the following mathematical model:

Yij = μ + αi + uj + eij

(3)

where

Yij = observed value for animal j receiving diet i; μ = overall mean; αi = fixed effect of diet i; uj = random effect of animal j; eij = residual error. A variance component (VC) covariance structure was used, as only one observation per variable was available per animal, and this structure assumes homogeneous variance across subjects. Normality and homoscedasticity of residuals were verified using Shapiro–Wilk and Levene’s tests, respectively. Treatment means were compared using Tukey’s test at a 5% significance level:

T = qα √(MSE/n)

where T is the minimum significant difference between treatment means, qα is the studentized range distribution value at significance level α, MSE is the mean square error from ANOVA, and n is the number of observations per treatment. This equation is now included to clarify the post hoc comparisons performed after analysis of variance.

3. Results

The diet containing S + CSC silage resulted in better animal performance, including higher TWG, ADG, FE, and FBW compared to the diet containing SS (p < 0.05) (

Table 3. Performance of lambs fed diets containing sugarcane silages added with cottonseed cake.

Item	Diets		SEM	p-Value
	SS	S + CSC
Total weight gain (kg)	6.41 b	8.11 a	0.51	0.0076
Average daily gain (g d−1)	160.23 b	202.88 a	12.61	0.0076
Initial body weight (kg)	26.10	26.33	3.44	0.5650
Final body weight (kg)	32.51 b	34.40 a	1.20	0.0054
Feed efficiency kg/kg DM	0.10 b	0.12 a	0.007	<0.0001

SS: solely sugarcane (control); S + CSC: sugarcane + 20% cottonseed cake (as fed basis); SEM: standard error of mean. Means followed by different letters, in same row, are significantly different by Tukey’s test (p < 0.05).

).

3.1. Performance Animal

Regarding intake parameters, there were no significant differences (p > 0.05) among treatments for dry matter (DM), organic matter (OM), neutral detergent fiber (NDF), acid detergent fiber (ADF), and total carbohydrates (TCs). However, diets containing the S + CSC silage resulted in significantly higher crude protein (CP), ether extract (EE), and total digestible nutrient (TDN) intakes (p < 0.05), indicating improved nutritional density with this silage combination (

Table 4. Nutrient intake and digestibility in lambs fed diets containing sugarcane silages added with cottonseed cake.

Item	Diets		SEM	p-Value
	SS	S + CSC
Nutrient intake (g/day)
Dry matter	1623.06	1621.75	71.14	0.4959
Organic matter	1592.42	1597.29	68.34	0.8447
Crude protein	243.32 b	265.50 a	10.59	0.0154
Ether extract	71.92 b	90.71 a	3.08	<0.0001
Neutral detergent fiberom	693.15	693.53	31.61	0.9886
Total carbohydrates	1313.65	1305.41	60.58	0.8695
Non-fiber carbohydrates	609.25	609.30	26.49	0.9981
Total digestible nutrients	1092.41 b	1207.60 a	39.52	0.0044
Nutrient digestibility (%)
Dry matter	61.04 b	64.75 a	0.45	<0.0001
Organic matter	63.71 b	67.19 a	0.45	<0.0001
Crude protein	74.97 b	80.06 a	0.57	<0.0001
Ether extract	77.97	78.00	1.25	0.9885
Neutral detergent fiberom	37.51 b	47.97 a	0.72	<0.0001
Non-fiber carbohydrates	87.40 a	82.23 b	0.77	<0.0001
Total carbohydrates	59.77 b	62.52 a	0.68	0.0004

SS: solely sugarcane (control); S + CSC: sugarcane + 20% cottonseed cake (as fed basis); SEM: standard error of mean. Means followed by different letters, in same row, are significantly different by Tukey’s test (p < 0.05).

).

3.2. Nutrient Intake and Digestibility

The inclusion of cottonseed cake in sugarcane silage improved the fermentation profile, as indicated by a higher pH, reduced ammonia nitrogen (N-NH₃) content, increased acetic acid levels, and lower ethanol concentrations (

Table 5. Fermentation profile of silages used in diets after 60 days of fermentation.

Item	Diets		SEM	p-Value
	SS	S + CSC
pH	3.20 b	3.67 a	0.06	<0.0001
WSC (g/kg−1)	48.80	46.0	0.15	0.6518
NH3-N g/kg−1 total N)	27.40 a	10.2 b	0.25	<0.0001
Lactic acid (g/kg−1)	30.70	37.70	0.61	0.3096
Acetic acid (g/kg−1)	13.70 b	20.0 a	0.22	0.0213
Ethanol (g/kg−1)	55.50 a	26.30 b	0.52	0.0024

SS: solely sugarcane (control); S + CSC: sugarcane + 20% cottonseed cake (as fed basis); SEM: standard error of mean; WSC: water-soluble carbohydrates; NH3-N: ammonia nitrogen. Means followed by different letters, in same row, are significantly different by Tukey’s test (p < 0.05).

).

3.3. Fermentation Profile of Silages

Ruminal fluid pH did not differ between treatments (p = 0.6865); however, N-NH3 concentration was higher in animals fed diets containing S + CSC silage (

Table 6. Ruminal parameters, ruminal volatile fatty acids concentration, and blood parameters in lambs fed diets containing sugarcane silages added with cottonseed cake.

Item	Diets		SEM	p-Value
	SS	S + CSC
	Ruminal parameters
pH	6.42	6.48	0.07	0.6865
N-NH3 (mg/dL−1)	7.34 b	7.77 a	0.30	0.0150
	Ruminal VFA (mmol/L−1)
Acetate	24.94 a	11.29 b	0.95	<0.0001
Propionate	6.52	5.78	0.53	0.3517
Butyrate	2.90	2.59	0.31	0.5548
Acetate/propionate ratio	4.23 a	2.02 b	0.16	<0.0001
Total VFA	34.34	19.66	0.59	0.4517
	Blood parameters (mg/dL−1)
Glucose	35.80 b	44.00 a	2.52	0.0554
Urea	19.50	22.80	1.63	0.2553

SS: solely sugarcane (control); S + CSC: sugarcane + 20% cottonseed cake (as fed basis); VFA: volatile fatty acids; SEM: standard error of mean. Means followed by different letters, in same row, are significantly different by Tukey’s test (p < 0.05).

). As for volatile fatty acids, differences were detected between treatments (p < 0.05), where acetate concentration in animals fed the diet containing S + CSC silage was significantly lower, with no difference for propionate and butyrate concentrations (Table 6). The acetate/propionate ratio was lower in animals fed the diet containing S + CSC silage (Table 6).

3.4. Ruminal Parameters, Ruminal VFA, and Blood Parameters

Blood glucose levels tended to differ between treatments (p = 0.0554; Table 6), with numerically higher values observed in animals fed the S + CSC silage compared to those receiving SS. Although this result did not reach the conventional threshold for statistical significance (p < 0.05), it suggests a potential trend that may warrant further investigation. Blood urea levels did not differ between treatments (Table 6).

3.5. Nitrogen Balance

Higher N intake was detected in the S + CSC treatment, with differences observed between treatments; similar results were found for urinary N and retained N (p < 0.05). An opposite trend was observed for fecal N, with levels also considered significantly different (p< 0.05) (

Table 7. Nitrogen balance in lambs fed diets containing sugarcane silages added with cottonseed cake.

Item	Treatments		SEM	p-Value
	SS	S + CSC
Intake N (g/day)	42.26 b	46.53 a	1.765	0.0142
Fecal N (g/day)	1.51 a	1.42 b	0.042	0.0363
Urine N (g/day)	1.05 b	1.44 a	0.075	0.0004
Retained N (g/day)	40.77 b	44.73 a	1.832	0.0277

SS: solely sugarcane (control); S + CSC: sugarcane + 20% cottonseed cake (as fed basis); SEM: standard error of mean. Means followed by different letters, in same row, are significantly different by Tukey’s test (p < 0.05).

).

4. Discussion

4.1. Fermentation Profile

The inclusion of cottonseed cake during ensiling altered the fermentation profile, increasing the pH to 3.67. This rise may be due to the dilution of water-soluble carbohydrates (WSCs) and the higher crude protein (CP) content, which can enhance the silage’s buffering capacity. Similar pH values above 3.60 were reported by Oliveira et al. [27]. Justino et al. [7] also observed average pH values around 3.70 in sorghum silage with cottonseed cake, while Santana et al. [8] found values near 3.90 in millet silage with cottonseed cake, both with and without microbial inoculants.

These moderately higher pH levels favor lactic acid bacteria (LAB) growth over yeasts, preventing alcoholic fermentation—commonly seen in untreated sugarcane silage. Sugarcane naturally has a high WSC content (300 g/kg), promoting strong acidification, as reflected by the lower pH in the control silage (Table 5). In contrast, the higher pH in S + CSC silage reflects reduced acidification due to WSC dilution from the 50% cottonseed cake inclusion (DM basis), which lowers the proportion of sugarcane. Both silages achieved adequate acidification, as NH₃-N levels remained well below the 10% threshold recommended by McDonald et al. [28]. According to the same authors, lower NH₃-N levels indicate reduced proteolysis and limited clostridial activity, leading to minimal butyric acid production.

4.2. Organic Acids

Among the organic acids produced, lactic acid (LA) was the most produced in the fermentation process, followed by acetic acid (AA) (Table 5). The activity of LAB during fermentation promotes the conversion of WSC into acids, mainly LA and AA [29]. Yeasts are undesirable during fermentation because they significantly increase DM losses in silages, resulting from the fermentation of WSC to ethanol and CO2 [29], with a reduction in the DM content of SS (Table 1). In contrast, the reduction in ethanol content in S + CSC silage may have occurred due to the residual unsaturated fatty acids (stearic and palmitic) present in the cottonseed cake, associated with the dilution of the WSC content, which may have interfered with yeasts, with reports of unstable membrane permeability and metabolic dysregulation in yeasts subjected to such stress [30].

Another factor that may also be associated with the reduction in ethanol is the smaller population of filamentous fungi and yeasts in the ensiled mass, possibly due to the deleterious effects of gossypol, the low water activity of the mixture, and the fermentation profile of the mixed silage [31]. Gossypol in cottonseed cake is a yellow polyphenolic compound characterized by one or more hydroxyl groups attached to aromatic rings. This compound has been reported to exert antioxidant and antifungal effects, which may help explain the lower ethanol content observed in S + CSC silage and contribute to its greater aerobic stability [32]. Although gossypol was not quantified in the present study, its potential role in modulating silage microbiota and fermentation characteristics should be interpreted with caution. Given that gossypol can be toxic at high levels, future studies should evaluate its concentration in cottonseed cake used for silage production and assess its implications for animal health in accordance with regulatory safety limits.

4.3. Performance Animal

Regarding the performance of animals fed the diet containing S + CSC silage, the mean values of TWG, ADG, and FE were significantly higher (Table 3) due to the greater digestibility of nutrients (Table 4). This is because, according to the NRC [13], this effect is more important than DMI in improving animal performance.

4.4. Nutrient Intake and Digestibility

Despite no differences in dry matter intake (DMI) between treatments, significant variations in digestibility and, consequently, animal performance were observed. The expected average daily gain (ADG) was achieved only with the S + CSC silage (Table 3), due to better preservation of its nutritional value (Table 1), reflected in the higher TDN of diets containing S + CSC (Table 2). Fernandez-Turren et al. [33] also reported increased nutrient intake in sheep fed PMR with alfalfa, regardless of whether the PMR energy source was starch or fiber.

A key factor for the improved performance of animals fed the S + CSC diet was the higher fiber digestibility (NDF), about 10% greater than in the SS diet (Table 4). This improvement is attributed to technologies such as PMR and TMR silages, which enhance nutrient preservation and digestibility [9]. Although fiber digestibility improved in the CSC treatment, NDF intake did not differ significantly, likely due to the physical characteristics of sugarcane, which limit voluntary intake.

Even though the diets were formulated to be isoenergetic and isoproteic (Table 2), differences were observed in CP, EE, and TDN intake (Table 4), resulting from the loss of soluble fractions in sugarcane silage (SS). As a result, only the S + CSC diet met the planned nutritional levels, leading to better animal performance. The lower TDN in the SS diet reduced overall nutrient intake, despite both diets exceeding the NRC’s recommended TDN intake (0.91 kg/day).

Actual DMI was higher than the calculated value (≈1.62 vs. 1.14 kg/day), likely due to greater weight gain during the final confinement phase and differences in breeds, climate, and feed typical of tropical regions. Although NRC [13] requirements are based on temperate conditions, they still offer a good reference.

Lastly, the use of PMR silage similar to TMR in fermentation profile improved digestibility through the action of fermentation acids, which hydrolyze proteins and fibers, releasing peptides, amino acids, pectin, and hemicellulose [9,34,35,36,37].

4.5. Ruminal Parameters, Ruminal VFA, and Blood Parameters

Animals fed diets with S + CSC silage showed higher ruminal NH₃-N concentrations (Table 6), likely due to increased protein degradability and overall diet digestibility, as reported by Kondo et al. [11] and Uyeno [38], even though nitrogen balance was not assessed in this study. Ruminal acetate levels and the acetate-to-propionate ratio also varied between diets (Table 6), influenced by dietary fiber content, which promotes acetate production [39]. The SS diet, richer in fiber and ethanol (Table 5), led to higher acetate levels, as ethanol is converted to acetate and methane in the rumen, often at the expense of propionate [40]. In contrast, the S + CSC diet showed reduced acetate levels due to lower fiber content, higher digestibility, and possible inhibition of fiber-degrading bacteria by lipids and bioactive compounds. This shift favored propionate production, enhancing energy efficiency without compromising animal performance [41].

The lower acetate/propionate ratio in the S + CSC group (Table 6) coincided with higher blood glucose levels, suggesting improved propionate utilization, supported by the higher TDNI and nutrient digestibility. Although ruminal propionate concentrations did not differ significantly, increased glucose levels can be attributed to (1) greater nutrient absorption and hepatic gluconeogenesis; (2) reduced ethanol-derived acetate production, minimizing metabolic competition; and (3) modulation of ruminal fermentation by unsaturated fatty acids and gossypol, which may have suppressed acetate-producing bacteria and enhanced propionate use for gluconeogenesis [40].

Propionate is a key gluconeogenic precursor, supplying energy for tissue deposition, particularly in animals fed S + CSC diets (Table 5) [41]. Improved nitrogen use was also observed in this group, reflected in lower fecal nitrogen losses and greater microbial protein synthesis (Table 7). However, the increase in urinary nitrogen excretion suggests a trade-off, with potential environmental impacts due to ammonia volatilization. Overall, higher glucose levels appear to result more from efficient propionate use and reduced ethanol production than from amino acid catabolism [41].

5. Conclusions

The inclusion of cottonseed cake in sugarcane silage improved the fermentative profile by reducing ethanol and ammonia nitrogen levels and increasing acetic acid concentrations, resulting in better nutrient preservation. These changes contributed to enhanced nutrient intake, digestibility, and productive performance in sheep. The findings support the use of cottonseed cake as a viable strategy to improve the quality of sugarcane silage and animal performance. Future studies should evaluate the long-term effects and monitor potential risks related to gossypol levels to ensure animal health and regulatory compliance.