Nutritional value of sorghum silages

. This study aimed to evaluate the nutritional characteristics and quality of silages of five sorghum genotypes, namely: Volumax, AG2005E, Qualimax, BRS610 and AG2501. The study was conducted at the Experimental Farm Unimontes, in the municipality of Janaúba, state of Minas Gerais. The treatments were distributed in a randomized block design with four replicates. Data were tested by analysis of variance and means were compared by Scott-Knott test at a significance level of 5%. The pH values ranged from 3.93 and 4.10. Genotypes differed in all studied nutritional characteristics (p < 0.05). For the acid detergent fiber, differences were detected between the genotypes, AG2501 (35.66%), Volumax (34.89%), AG2005E (34.53%), Qualimax (32.39%) and BRS610 (33.60%). The high participation of lignin was verified in all genotypes except for Volumax (4.14%). All silages were within the recommended range 50-65% for in vitro dry matter digestibility and classified as having good quality, however only the AG2005E genotype met nitrogen requirements for microbial fermentation, with 7.06% crude protein.


Introduction
In semi-arid regions, production of food for herds becomes the biggest challenge for producers due to climatic conditions.Thus, the problems caused by seasonality of forage can be minimized through food preservation practices.The lack of adapted cultivars that have desirable characteristics, such as high forage production, with high nutritional value, is the major difficulty faced by ranchers in the region in implementing the cultivation system.
Ensiling is the process of converting soluble carbohydrates of the forage into acids, mainly lactic and acetic acids, carried out by lactic acid bacteria under anaerobic conditions and is intended to maintain the quality and nutritional value of fresh forage.
The productive behavior of a forage used for silage production is one of the main characteristics to be used in the economic assessment of the activity.Estimating the nutritional value of forage is important, whether to allow for proper management of diets based on roughage or to guide the breeding or selection of forage, once the market provides a wide variety of sorghum cultivars (Moraes, Jobim, Silva, & Marquardt, 2013).For the production of good silage, it should be considered the percentage of grain in ensiled mass and high biomass productivity, participation of other fractions of the plant, and mainly the digestibility of neutral detergent fiber.Thus, studies comparing hybrids becomes important tools to contribute to breeding programs and to recommend hybrids or cultivars whose silages have the best production/nutritional value ratio (Antunes et al. 2007).
Therefore, the goal of the present study was to evaluate the nutritional value of silage of five sorghum genotypes grown in the winter.

Material and methods
The study was conducted at the experimental farm of the State University of Montes Claros -Unimontes, in the municipality of Janaúba, state of Minas Gerais, in the northern region of the state, in the Brazilian semi-arid region.
Five sorghum genotypes were analyzed, including three forage genotypes (Volumax, BRS 610 and Qualimax), a dual-purpose genotype (AG2005E) and the other for cutting and grazing (AG 2501).The planting was carried out in June 2014, in a randomized block design, with four replicates per genotype, totaling twenty plots with six rows of six meters each spaced 70 cm apart.Before planting, twenty soil samples were taken to a depth of 20 cm randomly throughout the experimental area using a Dutch auger.Soil portions were collected and mixed in a plastic container, comprising a composite sample; from this, a subsample was taken for physical and chemical analysis in the laboratory of the Department of Agricultural Sciences of Unimontes.
Based on the soil analysis, it was performed correction and fertilization of the soil.At planting, fertilization was made with 300 kg ha -1 of the formulation 04-30-10 (N-P-K).The topdressing was made at 35 days, using 60 kg ha -1 N, with urea as a source.For each plot, 20 seeds were planted and, after emergence, seedlings were thinned to adjust the number of plants per linear meter.
Harvest was held in November, totaling an experimental period of 119 days.We disregarded 1 m from the ends of each row and the two lateral rows of each plot (borders).The two middle rows of each plot were used for ensiling when each genotype showed dry matter content between 30 and 35%.For ensiling, we used laboratory silos made of PVC pipes 100 mm in diameter, 500 mm length, with average density 600 kg m -3 .The forage was minced using a stationary shredder and pressed with wooden socket.Silos were sealed immediately after filling, with PVC caps fitted with Bunsen valves and sealed with tape, and were weighed before and after ensiling.
There were made four replicates per treatment and three replicates per plot; a total of 60 silos were filled, which were opened after 56 days of ensiling.Nutritional assessment of silages was carried out at the Food Analysis Laboratory of Unimontes.
Four replicates per treatment and three replicates per plot were made and a total of sixty (60) silos were made, which were opened after 56 days of ensiling.The nutritional evaluation of the silages was carried out at the Laboratory of Food Analysis of Unimontes.
Upon silo opening, the material was homogenized and part of the ensiled material was placed in paper bags, weighed and pre-dried in a forced ventilation oven at 55°C for 72 hours or until constant weight.Pre-dried samples were ground in a stationary mill with a 1 mm mesh sieve and then placed in glass vials with screw cap and identified for food chemical analysis.The variables analyzed were: dry matter (DM), ash, crude protein (CP), neutral detergent fiber (NDF), acid detergent fiber (ADF), cellulose, hemi cellulose and lignin, according methodologies described by Detmann et al. (2012).
The evaluation of in vitro dry matter (DM) digestibility was performed in samples pre-dried in a forced ventilation oven at 55°C for 72 hours, and ground with 5 mm mesh sieve.IVDMD was determined according to the method described by Tilley and Terry (1963), modified by Holden (1999), through the use of in vitro incubator Tecnal ® (TE-150), with modification of the bag (7.5 x 7.5 cm) material, which was made using nonwoven fabric (TNT -100 g m -2 ) according to Casali et al. (2014).The rumen fluid necessary for the evaluation was collected from two adult crossbred steers, castrated, fitted with rumen cannula, approximately 400 kg body weight.Animals were housed in a pen, fed with forage (sorghum silage), concentrate, mineral salt and water ad libitum for 15 days prior to the collection.Data were tested by analysis of variance using the software Sisvar (System for analysis of variance) and means were compared by Scott-Knott grouping test at a significance level of 5%, according to the following statistical model a Equation 1 In which: Yĳ= value related to the observation of hybrid j in blocki μ = overall mean; ti = effect of treatment on hybrid j (j = 1, 2, 3, 4, 5); bj = effect of blocki (i = 1, 2, 3, 4); ℮ĳ = experimental error associated with the experiment.

Results and discussion
The results on the quality of silage of five sorghum genotypes are listed in Table 1.The mean values of pH ranged between 3.93 and 4.10, and showed no significant differences (p > 0.05).In this study, all silages had pH near 4.0 and can be classified as excellent.Thus, the obtained pH values indicated that all treatments had sufficient availability of soluble carbohydrates for adequate fermentation.In relation to NH 3 -N/TN content, all silages can be considered as of good quality, because all the values were below 10% (Table 1).For this characteristic, differences (p < 0.05) were detected between the genotypes AG2501 (8.53), BRS610 (8.00) and AG2005E (7.74), which did not differ to each other and were superior to Qualimax (7.74) and Volumax (5.78), which were similar, at 5% probability.The NH 3 -N/TN content values are related to DM content.Although AG2005E, BRS610 and AG2501 presented high DM content, the NH 3 content values were higher, which indicates that there must have been some experimental error at the time of ensiling that favored a clostridial fermentation and consequently high proteolysis, raising the ammonia nitrogen content.
Variations occur depending on the type of stem, dry or wet.All tested genotypes have wet stem and very similar concentration of carbohydrates.BRS610 is a hybrid obtained from crossing BR601, of wet stem, with BR700, of dry stem.Thus, the genes of stem trait should have had dominance and produced higher content of soluble carbohydrates, providing more substrate to lactic acid bacteria, increasing the lactic acid content.With respect to AG2501, inferiority in lactic acid content seems to be associated with the characteristics of this genotype, since there are changes in the patterns of use of soluble carbohydrates for production of these compounds.
Table 1 also shows the percentages of acetic acid, in which AG2501 (2.28) surpassed all others; while Volumax (1.48), AG2005E (1.49) and Qualimax (1.50) did not differ to each other, but exceeded the BRS610 (1.10).AG2501, for cutting and direct grazing, has its point of harvest at 50 days; and as the cut for silage was late, at 80 days, increased the DM content (34.99%) by changing the profile of soluble carbohydrates and hence the proliferation of hetero fermentative bacteria, raising the content of acetic acid.BRS610 was inferior to others, because it is an early hybrid harvested at the optimal time, allowing a good lactic fermentation and, consequently, minimizing the presence of hetero fermentative bacteria, which kept low the acetic acid content.A high acetic acid content may restrict the lactic fermentation, as observed with genotype AG2501, wherein the percentage of acetic acid (2.28) exceeded the others and consequently lowered the lactic acid content (4.40), which was lower than the others.The opposite was observed with BRS610, where the acetic acid content was 1.10%, lower than the others, and the lactic acid content was 6.82%, exceeding all the others.
The mean values of the nutritional characteristics of five sorghum genotypes are listed in Table 2.In relation to dry matter, genotypes AG2501 (34.99),BRS610 (33.84) and AG2005E (33.33) were similar to each other (p > 0.01) and superior to Volumax (31.01) and Qualimax (31.75), which were not different to each other (p > 0.01).
AG2005E stood out as it is a dual-purpose genotype, presenting higher percentage of panicle that contributed to the higher dry matter content.As for BRS610, the superiority in the dry matter content is because it is an early hybrid, which completes early the vegetative cycle and thus was at a more advanced stage of maturity at the time of cutting, raising the dry matter content.
AG2501 is a hybrid for cutting and grazing, has greater height, which raises the stem percentage (81.44%).Volumax and Qualimax were inferior to the others because they are late genotypes, but had normal content for ensiling at the time they were cut.
The optimal dry matter content for ensiling is estimated at 30-35% (Dias et al., 2010) to avoid losses by the formation of effluents and biological processes that produce gas, water and heat, as well as provide adequate lactic fermentation to maintain the nutritive value of the silage.Tolentino et al. (2016) reported values of 37.12% for sorghum of Volumax and 39.48% DM for BRS610, with values higher than those found in the present experiment for the same genotypes.
A food and/or diet should contain at least 7% CP to provide sufficient nitrogen for effective microbial fermentation in the rumen (Church, 1988).The percentage of crude protein is not modified with the ensiling process, although different proportions of nitrogen fractions may be altered.It is known that fermentation causes changes in the composition of nitrogen fractions, reducing the true protein content and increasing the content of free amino acid, or products from the breakdown of these of amino acids, including ammonia, CO 2 and amines (Ohshima & McDonald, 1979).
Thus, the low nitrogen content of silages is a limiting factor for the use of forage of these genotypes mainly for animals with high nutritional requirements.Tolentino et al. (2016) found CP content above 9.06% when evaluated silages of twenty-four different genotypes of sorghum, values higher than those presented in this study.
Values of neutral detergent fiber (NDF) content of the silages did not differ to each other (p > 0.05), ranging from 59.08 to 61.85% (Table 2).For the microbial balance in the rumen, it is important to have a minimum amount of fiber to maintain the fermentation and stimulate salivation, which are indispensable for the fermentation process and prevent metabolic disorders.NDF and ADF contents are indicative of the amount of fiber in the forage; NDF is related to the amount of fiber that is in the roughage, while the FDA, to the amount of less digestible fiber.In this sense, the lower the content, the better the quality of the silage produced and the higher the DM intake by the animal (Santos, Galvão, Silva, Miranda, & Finger, 2010).NDF values above 60% are negatively correlated with dry mater intake by the animal, as for the ADF fraction, high content values hinder the fragmentation of food and digestion by ruminal bacteria.
The ADF is related to the forage digestibility, because it contains the higher proportion of lignin, which is the completely indigestible fiber fraction, thus indicating its indigestibility.The ADF content should not exceed 30%, thus high levels, as found in this study, hinder the absorption of nutrients.
Values of hemi cellulose content of the silages were not different to each other (p > 0.05) and ranged from 25.48 to 27.36%.In forage at later stages, hemicellulose appears more associated with lignin by covalent bonds than others polysaccharides, making them unavailable.
For the cellulose content found in silages, in percentage of DM, there were differences between genotypes (p > 0.05).Volumax (27.75), AG2005E (27.23) andAG2501 (28.45) were similar to each other (p < 0.05) and superior to Qualimax (26.05) and BRS610 (25.69) (Table2).These values are directly related to ADF values, since cellulose is an important component of this fiber fraction.
As it is the main component of ADF, the superiority in the cellulose content of AG2501 and Volumax is similarly explained by their height, 2.40 and 1.91, respectively, while inAG2005E, the result

Table 1 .
Mean values of pH, ammonia nitrogen/total nitrogen (NH 3 /TN), lactic acid, acetic acid and in vitro dry matter (DM) digestibility of silage of five sorghum genotypes grown in the winter.
Mean values followed by different lowercase letters, in the same row, are significantly different (p < 0.05). 1 On a drymatterbasis.