1. Introduction
The pursuit of alternative feed sources for ruminant nutrition is crucial, not only for fostering more sustainable animal production from both environmental and economic perspectives, but also for ensuring enhanced human food security [
1]. In semi-arid regions characterized by extended drought periods, alternative feed sources are essential due to the prevalence of extensive production systems. These systems experience a reduction in both the quantity and quality of forage during drier periods, resulting in decreased productivity and performance [
2].
Supplementation with concentrates during the dry season is generally not a cost-effective practice due to the high feeding costs associated with the use of expensive ingredients such as ground corn and soybean meal, especially during periods of forage scarcity in these regions.
Using low-cost feeds that are adaptable to the drought periods characteristic of semi-arid regions, with productive and nutritional potential, is an alternative with sustainable viability. From this perspective, forage palm is widely used as animal feed in semi-arid regions due to their adaptability to these environments [
3,
4]. Despite presenting interesting nutritional characteristics for livestock feed, forage cactus, such as spineless cactus (
Nopalea cochenillifera Salm Dyck), has low levels of dry matter (DM, 6.37%), crude protein (CP, 5.42%), and neutral detergent fiber (NDF, 22.05%) [
5].
Spineless cactus can be fed to animals fresh, almost immediately after being harvested [
6] or in the form of silage [
4], constituting an important source of energy, water [
6,
7] and non-fiber carbohydrates (NFCs), which include galactose, arabinose, xylose, fructose, and, especially, pectin [
8]. However, due to its low nutritional content of CP and NDF, this feed source does not fully satisfy the nutritional requirements of animals. Therefore, it is essential to explore alternatives to enhance spineless cactus silage, such as the use of forage consortia with other feed sources adapted to the semi-arid region for silage production. As Matias et al. [
4] point out, the use of protein ingredients such as maniçoba (14.5% CP) to ensile tuna improves the nutritional quality and fermentative characteristics of the silage.
Some of the main legumes used in the Brazilian semi-arid region are gliricidia (
Gliricidia sepium), catingueira (
Caesalpinia bracteosa), “sabiá” (
Mimosa caesalpiniifolia Benth) and “catanduva” (
Piptadenia moniliformis Benth), due to important characteristics for animal production, such as perenniality, high CP content (>16%), and average rumen dry matter degradability of 50–60% DM [
9,
10].
Considering the various possibilities of forage associations to be used in silage production, gliricidia emerges as an alternative because it is a legume adapted to semi-arid conditions and has multiple uses in agro-silvicultural systems, as a protein source in diet supplementation, at low cost, for ruminants [
11].
As described by Costa et al. [
12], gliricidia has desirable characteristics in the composition of its leaves, with average contents of 23.1% DM, 24.1% CP, and 38.8% NDF. The CP content stands out for presenting above 7%, which is the minimum necessary to ensure the balance of rumen microorganisms, promoting the correct functioning of the rumen [
13]. According to Lemos et al. [
14], the use of gliricidia hay promotes an approximately 50% increase in DM intake and average daily gain in feedlot lambs.
In this context, gliricidia hay can serve as a complement to spineless cactus in silage production, functioning as an absorbent additive that enhances the DM, CP, and NDF content of the resulting silage. Furthermore, it will improve the fermentative and nutritional characteristics of the silage, thereby increasing the overall efficiency of the process [
15].
To achieve optimal productivity efficiency, a diet should be formulated to contain the ideal nutritional characteristics that promote maximum productive efficiency [
16]. For this reason, ground corn, cottonseed cake, and soybean meal are commonly used as the primary ingredients in concentrate production for their nutritive characteristics [
17,
18]. However, gliricidia hay emerges as an excellent feed alternative due to its low cost and its ability to replace expensive ingredients in the production of high-quality diets. Therefore, further research is required to determine the nutritional potential of these feed sources [
11].
In light of the above, and with the aim of producing silages from plant species adapted to semi-arid regions that maintain the nutritional quality necessary to meet the needs of ruminants, we hypothesize that the inclusion of up to 20% gliricidia hay (% fresh matter—FM) can replace ground corn and cottonseed cake in the production of total mixed ration silages based on spineless cactus. This inclusion is expected to improve the fermentation process while preserving the ideal nutritional quality of the final product. Therefore, this study was conducted with the objective of evaluating the chemical composition, fermentation profile and losses, microbial population, and aerobic stability of total mixed ration silages based on spineless cactus, with increasing levels of gliricidia hay replacing ground corn and cottonseed cake.
2. Materials and Methods
2.1. Location
The cultivation and harvesting of forages, along with the making of total mixed ration silages, were conducted in the Instituto Federal de Educação, Ciência e Tecnologia Baiano, Campus Santa Inês, Bahia, Brazil. Based on the Köppen–Geiger classification, the climate type of the municipality is tropical savannah (Aw). Temperatures usually range from 20 °C to 24 °C throughout the year, but they can rarely fall to 12 °C or can increase as high as 34 °C. The average annual rainfall of ~853 mm.
After filling, sealing, and weighing, the silos were taken and stored in the Forage Laboratory at the Experimental Farm of São Gonçalo dos Campos of the School of Veterinary Medicine and Animal Science of the Universidade Federal da Bahia. The farm is located at Km 174 of the BR 101 highway, Mercês city, Municipality of São Gonçalo dos Campos—BA, at the coordinates 12°23′57.5″ south latitude and 38°52′44.6″ west longitude, located at a distance of 108 km from Salvador.
2.2. Experimental Design and Spineless Cactus Silage Making
Twenty-five experimental silos (five treatments and five replications) were used in a completely randomized design. The cottonseed cake and the other concentrated ingredients (ground corn and gliricidia hay) were used in the formulation of the total mixed rations (
Table 1).
The spineless cactus (forage cactus cv. “Miúda”) was harvested manually and processed in a stationary forage chopper. The chopper was adjusted to cut a particle size of approximately 2 cm for the spineless cactus. As for the gliricidia hay, the material for production was harvested from an established plantation of gliricidia approximately 6 months earlier on a farm in Brejões-BA, which is around 45 km from Santa Inês. The material was taken to the IF Baiano, Santa Inês Campus, where it was produced with dehydration by sunlight. It was then ground to cut a particle size of approximately 1 cm and stored in an airy shed.
Treatments referred to the gliricidia hay inclusion levels (0, 5, 10, 15, and 20%, FM). At the time of ensiling, spineless cactus was mixed with the ingredients in the respective proportions for each treatment until the mixtures became homogeneous. Thus, the material was manually mixed and promptly compacted with cement punches to achieve a final packing density of 700 kg of fresh matter/m
3 in each experimental silo. The experimental diets, in the form of total mixed rations, were formulated to meet the sheep’s requirements for an average daily gain of 200 g as proposed by the NRC [
19].
Table 2 describes the ingredient proportions of the treatments.
The silages were made in 25 experimental polyvinyl chloride silos, which measured 10 cm in diameter and 45 cm in height. All silos were adapted with a Bunsen valve adapted to the lid for the elimination of gases resulting from fermentation.
At the bottom of each mini-silo, 0.6 kg of dry sand was added and covered with non-woven fabric to separate it from the chopped forage (approximately 2.5 kg) and to capture and measure effluents. At the end of this process, all silos were closed using adhesive tape, weighed, and stored in a place (covered, dry, and ventilated) at room temperature until the opening 120 days after ensiling.
2.3. Fermentation Losses
Fermentative losses in the form of gases (GLs), effluents (ELs), and total (TLs) were quantified in the spineless cactus silage by weight difference through the following equations described by Jobim et al. [
20]:
where: GL = gas losses (% dry matter); SWC = weight of full silo at closure (kg); SWO = weight of full silo at opening (kg); FMC = forage mass at silo closure (kg); and DMCC = forage dry matter content at silo closure (%).
where: EL = effluent losses (kg/ton FM); ESWO = weight of empty silo + sand at opening (kg); SW = weight of empty silo (kg); ESWC = weight of empty silo + sand at closure (kg); and FMC = forage mass at silo closure (kg).
where: TL = total loss of DM (%); DMi = initial DM. Weight of silo after filling—weight of the empty set, without forage, before filling (dry tare) × DM content of the forage in the silage and DMf = final DM. Weight of the silo before opening—weight of the empty set, without the forage, after opening the silos (wet tare) × DM content of the forage at the opening.
where: DMR = dry matter recovery rate (%); FMO = forage mass at silo opening (kg); DMO = DM at the opening (%); FMC = forage mass at silo closure (kg); and DMC = forage DM content at silo closure (%).
2.4. Sample Collection and Assessment of the Chemical Composition
After evaluating the fermentation losses, spineless cactus silage samples were collected from each silo and submitted to pre-dried in a forced ventilation oven (55 °C for 72 h) (INCT G-001/2) and processed in a Willey mill (Tecnal
®, Piracicaba, São Paulo, Brazil) with a 1 mm sieve and analyzed for evaluations of DM (G-003/1), ash (INCT-CA M-001/2), CP (INCT-CA N-001/2), ether extract (EE; INCT-CA G-004/1), NDF (INCT F-001/2), and acid detergent fiber (ADF: INCT F-003/2) contents according to Detmann et al. [
21]. Organic matter was estimated using the equation: 100—ash content. Non-fibrous carbohydrates (NFC) and total carbohydrates (TC) contents were calculated through the Weiss [
22] and Sniffen et al. [
23] equations, respectively.
2.5. Fermentation Profile of Total Mixed Ration Silages
The pH values were measured in all silos using a benchtop pH meter (KASVI,
® São José dos Pinhais, Paraná, Brazil) according to the methodology proposed by Bolsen et al. [
24]. The ammoniacal nitrogen (NH
3-N) of the silages was determined using distillation with potassium hydroxide (KOH) 2N, as described by Fenner [
25].
Organic acids (i.e., lactic, acetic, propionic, and butyric acids) were determined using the methodology described by Canale et al. [
26] with the aid of a Shimadzu high-performance liquid chromatography (HPLC) system equipped with an SPD-M20A—UV DETECTOR (SHIMADZU) and an AMINEX HP87-H column (Bio-Rad Laboratories Ltd., São Paulo, Brazil) (30 cm × 4.5 mm in diameter). These analyses were carried out with a column flow rate of 0.6 mL min
−1, a column pressure of 6.0 MPa, a temperature of 39 °C for 30 min, and an injection volume of 20 µL.
Before the analysis, the samples were processed and conserved in 20% metaphosphoric acid. Then, they were kept at −20 °C until further analysis. Although all the acids were evaluated, the propionic and butyric acid concentrations were below the measurable levels in all silage samples. Thus, the results of these acids could not be stated in the results table.
2.6. Microbiological Analyses of Silage
The microbial populations were quantified after opening of the silos, with the aid of selective culture media for each microbial group evaluated. Thus, the counts of lactic acid bacteria (LAB) were assessed using MRS agar (Man, Rogosa, and Sharpe; KASVI
® São José dos Pinhais, Paraná State, Brazil) with the addition of 0.1% acetic acid. Molds and yeasts were quantified using potato dextrose agar (PDA, KASVI
® São José dos Pinhais, Paraná, Brazil) acidified with 1% tartaric acid (weight/volume) following the methodology described by González and Rodríguez [
27].
The microbial populations were quantified from 10 g of a sample of each silo, which was homogenized in 90 mL of sterile distilled water for 1 min, considered as dilution of 1/10. Afterward, successive dilutions were performed (10−1 to 10−6) for each group and placed in sterile disposable Petri dishes. Then, the selective culture medium was included in each plate, which was incubated in B.O.D (Biochemical Oxygen Demand) (lactic acid bacteria—LAB: 37 °C for 48 h; mold and yeasts: 37 °C for 72 h).
After the incubation period in B.O.D. at the above-mentioned times, the colonies were counted. The morphological characteristics of the colonies helped differentiate molds and yeasts.
Dishes with values between 30 and 300 colony-forming units (CFU) were considered countable and used to evaluate the microbial populations for each group. After the counting, all values were converted to a logarithmic scale (base log10). The number of mold and yeast colonies was lower than the values that should be considered, and these data are not included in the results section.
2.7. Assessment of Aerobic Stability
After 120 days of ensiling, the aerobic stability (expressed in hours) was assessed by evaluating the surface and internal temperatures of the spineless cactus silages submitted to air exposure.
Aerobic stability was evaluated by returning 1.0 kg of the silage samples to their respective experimental silos, without compaction and lids, and maintained in a closed environment with controlled temperature (25 °C).
The temperatures (internal of the silages and the environment) were evaluated every 2 h. The internal temperatures were recorded using a skewer thermometer (KASVI
® São José dos Pinhais, Paraná, Brazil), which was placed in the geometric center of the silage mass. In contrast, the ambient temperature was measured with a digital laser thermometer. The maximum and minimum temperatures (°C), thermal amplitude, and forage losses (% FM) were evaluated during the aerobic stability evaluation. In addition, the beginning of deterioration was considered when the internal temperature of the silages reached 2 °C above the ambient temperature [
28].
2.8. Statistical Analysis
The experiment was conducted in a completely randomized design with 5 treatments and 5 replications, totaling 25 experimental units. The following statistical model was used:
where: Ŷ
ij = observed value of the dependent variable; μ = overall average; T
i = fixed effect of gliricidia hay inclusion; and Ɛ
ij = experimental random error associated with each assumption observation NID~(0, σ
2).
The data were submitted for analysis of variance using the Bartlett test and regression. The degrees of freedom were decomposed by orthogonal contrasts in linear and quadratic effects, according to the inclusion levels of gliricidia hay in the spineless cactus silage. The significance of the regressions was determined at the 5% probability level for type-I error, using the PROC MIXED of the SAS 9.4 statistical package program.