Chemical composition and nutritive value of South African sorghum varieties as feed for broiler chickens

Copyright resides with the authors in terms of the Creative Commons Attribution 2.5 South African Licence. See: http://creativecommons.org/licenses/by/2.5/za Condition of use: The user may copy, distribute, transmit and adapt the work, but must recognise the authors and the South African Journal of Animal Science. ________________________________________________________________________ Abstract Sorghum (Sorghum bicolor (L.) Moench) is the fifth most important grain crop after wheat, rice, maize and barley. It is cultivated for food and feed in America, Asia, Australia and Africa. Newly developed sorghum varieties should be evaluated for their suitability as food and feed. The physical characteristics and proximate composition, total phenolic content, mineral content, amino acid profile and digestibility, and true metabolizable energy of four sorghum varieties were studied. The condensed tannin sorghum varieties PAN8625 and NS5511 had higher total phenolic content and antioxidant activity than the non-tannin varieties PAN8816 and PAN8906. Crude protein content diverged narrowly across the varieties, ranging from 81.2 to 95.4 g/kg DM. Starch and gross energy contents also differed, though varieties that were used had similar total and individual mineral contents. The threonine, leucine, phenylalanine, valine, proline and alanine contents of PAN8625 and PAN8906 were higher than those of NS5511 and PAN8816, which had similar contents. Amino acid digestibility and metabolizable energy of the tannin varieties were generally lower than those of the non-tannin ones. These results indicate that chemical and nutrient composition of sorghum varieties should be considered when selecting for broiler chicken feeding. ______________________________________________________________________________________


Introduction
The use of maize as the main energy source in poultry diets is subjected to the age-old competition between animals and humans.The grain is increasingly being used to produce ethanol, particularly in North America.Many areas in the world are not self-sufficient in grain production to meet human and animal needs and have depended on imports from North America.Maize grain prices have therefore remained high and will probably increase in future.Therefore, alternative, accessible and cost-effective ingredients are needed that can provide the same nutrients as maize.Sorghum (Sorghum bicolor) has been identified as a suitable replacement for maize (Medugu et al., 2010).It is abundant in most parts of Africa and Asia, and is among the main cereal crops that are used as food and feed.The adaptive agronomic characteristics of sorghum make it suitable for cultivation in different environmental conditions.
Sorghum varieties have been used for many years for poultry and human consumption, especially in Africa and Asia, and have been shown to vary in physical and chemical characteristics (Neucere & Sumrell, 1980;Sedghi et al., 2011).While some authors have reported anti-nutritional factors in sorghum grain that have adverse effects on the utilization of sorghum protein and metabolizable energy by poultry, such as tannins (Gous et al., 1982;Boren & Waniska, 1992), others have stated that these factors do not always reduce the performance of poultry (Nyachoti et al., 1997).This indicates the possibility of other factors contributing to the poor feeding value of certain sorghum varieties.In human beings, sorghum that is consumed as refined or whole grain has proved to have potential health benefits (Mabelebele & Iji, 2013).The nutritive values of several South African sorghum varieties in broiler diets have not yet been reported.Therefore, the objective of the current study was to determine the physical characteristics and nutritive value of four sorghum varieties as potential alternative feed ingredients for poultry.

Materials and Methods
Four commercial hybrid sorghum grain varieties grown in a controlled field trial at Pannar Research Services (Pty) Ltd, Klerksdorp, South Africa, were used in this study.The mean temperatures around the area are above 29 ºC in summer and below 21 ºC in winter.It lies at latitude 26.55'8 S and longitude 26.40'0 E. These varieties were grown in the 2010 -2011 season under dry land conditions that receive a mean annual rainfall of less than 482 mm.
The varieties that were selected for their outstanding yield performance, agronomic characteristics, preference by farmers and tannin contents were Pannar seed hybrids (PAN8625, PAN8816, PAN8906) and a national seeds hybrid (NS5511).The grains were field dried and harvested at less than 14% moisture content.Five kilograms of each variety were cleaned to remove broken and foreign materials prior to laboratory analysis.
The thousand-kernel weight (TKW) was determined by weighing 1000 sound kernels as a representative sample.The colour of the grain was observed visually and HunterLab L (lightness), a (redness) and b (green) colour values were measured with a Konica Minolta CR-400 c camera (Konica Minolta, Sakai, Osaka, Japan).A total of 20 sound kernels of each variety were cut in half with a blade, and kernel texture was assessed for the proportion of corneous and floury endosperms.The dissected kernels were viewed with a Leica M216 stereomicroscope equipped with a Leica DF C450 digital camera (Leica Microsystems, Heerbrugg, Switzerland).The outer layer (pericarp) of kernels of the four sorghum varieties was scratched with a scalpel and viewed under the light microscope to check for the presence or absence of a pigmented testa.The bleach test was used to detect sorghum grains with a pigmented testa (Taylor, 2001).The total phenolic concentration of milled grain of each sorghum variety was determined with the Folin-Ciocalteu assay described by Singleton & Rossi (1965).The vanillin-HCL assay of Price et al. (1978) was used to measure the condensed tannin content.Absorbance was read at 500 nm and catechin was used as a standard.
Proximate analysis for moisture, ash, crude protein (N × 6.25), fat and starch was carried out according to the standard methods of AOAC (2005).Ground samples of each variety were oven-dried and weighed.The sample, in a crucible, was ashed in a muffle furnace at 550 ºC for 6 hours.The ash was aciddigested by adding 1 mL 55% (v/v) HNO 3 .After cooling, calcium, magnesium, manganese, zinc, iron, sodium, potassium, copper, sulphur and phosphorus concentrations were determined by AOAC method 6.1.2(2005) using inductively coupled plasma spectroscopy.Neutral detergent fibre (NDF) and acid detergent fibre (ADF) were established according to the methods of Van Soest et al. (1991).The gross energy contents of the milled samples and excreta were determined with adiabatic bomb calorimetry (Gallenkamp, Autobomb, London, UK).
The modified intubation assay of McNab & Blair (1988) was used to determine the apparent metabolizable energy (AME) and true metabolizable energy (TME) contents of the cultivars.A total of 48 adult male laying-type birds were used in the evaluation, having been fasted for 48 h prior to being fed the samples.The roosters were kept in battery cages and were assigned to four treatments (sorghum varieties), replicated three times with four birds per replicate.
The birds were given 50 g glucose dissolved in 50 mL water by intubation after 24 h fasting to ensure an adequate energy supply during the fasting period.Twelve birds per variety were then given 50 g milled sorghum by intubation.Excreta were collected in trays beneath each cage over the following 48 h, dried at 70 ºC in an oven for 48 h and then weighed.The milled sorghum and excreta samples were hydrolysed with 6 N HCl at 110 ºC for 24 h (AOAC, 2005) and analysed for amino acids using Shimadzu HLPC reversephase high-pressure liquid chromatography (Shimadzu LC-10 AD, Shimadzu Corporation, Kyoto, Japan) with Shimmpack amino-Na-type column, 10 cm × 6.0 mm (Shimadzu Corporation).Samples were post column derivatized with o-phaldialdehde and detected with the Shimadzu RF-10A fluorescent detector (Shimadzu Corporation).Data were recorded and integrated using a Chromatopack CR7-A integrator model (Shimadzu Corporation).AME, TME and amino acid digestibility values were calculated with the following formulae with appropriate corrections for differences in dry matter (DM) content: TME (MJ/kg) = [(feed intake × GE diet ) -(excreta energy -excreta energy of the fasted bird)] / feed intake 3.
Amino acid digestibility = [AA consumed -AA excreted / AA consumed ] × 100 Samples were collected in triplicate from each variety to determine the chemical composition and nutritive value of the grains.Data were analysed using a one-way analysis of variance of SAS statistical analysis software (SAS, 2008).When ANOVA indicated a significant P-value, means were separated using Duncan`s multiple range test.Pearson's correlation coefficient was used to decide whether there were linear relationships between Hunter LAB, texture, antioxidant activity, total phenolic and condensed tannins.

Results and Discussion
The kernel characteristics of the grains are presented in Table 1.The TKW ranged from 33 to 28 g, with PAN8906 having the highest (P <0.05) TKW.The visual kernel colour of all the grains was brown, except for PAN8906, which was red.There were significant (P <0.05) differences in HunterLab values among the varieties.The HunterLab values ranged from 41.6 (lightest) to 38.2 (darkest).PAN8816 was harder (P <0.05) than other varieties (Table 1).PAN8625 and NS5511 had pigmented testa, whereas this was absent in the other two varieties (Table 2).The varieties that had pigmented testa tested positive for the bleach reaction, while the varieties without a pigmented testa tested negative.Thus, in the current study, the bleach reaction and the scratch test results indicate that PAN8625 and NS5511 contained condensed tannins, while varieties PAN8816 and PAN8906 were tannin-free.Indeed, PAN8625 and NS5511 had higher condensed tannin concentrations than PAN8816 and PAN8906 (Table 2).The tannin varieties in the present study were reported to have higher total phenolic concentrations than the non-tannin varieties.These results are similar to the reports of Awika et al. (2005), Dykes et al. (2005) and Chiremba et al. (2009).The correlation analysis suggests that the condensed tannins in the tannin varieties PAN8625 and NS5511 contributed to their higher antioxidant activity relative to the non-tannin varieties PAN8816 and PAN8906 (Table 3).Many authors working with sorghum varieties containing different phenolic types have reported similar results and reached similar conclusions (Awika et al., 2003;2005;Dykes et al., 2005;Chiremba et al., 2009).The proximate composition of the four sorghum varieties is presented in Table 4.The ash value of PAN8816 was lower (P <0.05) than values for PAN8625, NS5511 and PAN8906.Fat content varied among the varieties, ranging from 27 to 37 g/kg DM.These fat contents were slightly lower than the sorghum cultivars evaluated by Gassem & Osman (2003), which ranged from 35.8 to 44.7 g/kg DM, although they were similar to those reported by Basahy (1995) and Abdel-Rahman (2000).The ADFN and NDFN contents diverged, with ADFN varying to a greater extent (47 to 82 g/kg DM) than the NDFN contents (117 to 148 g/kg DM).The tannin varieties had higher ADFN and NDFN concentrations than the non-tannin ones (76 and 143 vs. 53 and 118 g/kg DM, respectively).The ADFN and NDFN concentrations of the tannin sorghum varieties evaluated by Parnian et al. (2013) were considerably higher (139 and 224 g/kg DM, respectively) than those measured in the current study.The contribution of polyphenols to the lignin fraction of the dietary fibre is usually responsible for the higher values of dietary fibre in the tannin varieties.a,b,c,d Means in the same column with different superscripts are significantly different (P <0.05).SEM: standard error of mean.

Table 1 Kernel characteristics of four sorghum varieties
The crude protein level ranged from 81 to 95 g/kg DM, averaging 89 g/kg DM.These values were similar to those of Gualtleri & Rapaccini (1990), Rooney & Serna-Saldivar (2000) and Issa (2009) who reported crude protein contents of sorghum ranging from 60 to 100 g/kg DM.The starch content varied narrowly between 503 and 547 g/kg DM.The highest starch content (542 g/kg DM) was observed in PAN8625.This value was lower than the average starch content reported for sorghum generally, which ranges from 560 to 730 g/kg DM (Geleta et al., 2005;Ragaee et al., 2006;Boudries et al., 2009).The variations in chemical and proximate composition could be owing to differences in genotype and growing conditions such as geographical and seasonal variations, climatic conditions and soil characteristics.The mineral concentrations varied among the four varieties (Table 5).The non-tannin variety, PAN8816, generally had a lower mineral concentration than the other varieties.In the current study PAN8906 and PAN8625 had higher mineral concentration than the other varieties.However, the calcium level obtained in the current study was lower than that reported by Shegro et al. (2012), although Samia et al. (2005) also recorded low values, similar to those in the current study.The highest value for Na, namely 40 mg/kg DM, in this study is lower than that reported by Awadelkareem et al. (2009) and Shegro et al. (2012), who reported values of 63.0 -70.0 and 58.3 -61.8 mg/kg DM, respectively.
The amino acid composition of the four sorghum varieties varied little among varieties (Table 6).There was no difference (P >0.05) between varieties in the concentration of histidine, lysine, methionine, cysteine or serine.The amino acid digestibility is presented in Table 7.This was generally lower in PAN8625 and NS5511 than in the non-tannin varieties PAN8816 and PAN8906.NS5511, which had high condensed tannin levels, yielded a low amino acid digestibility for arginine, histidine, phenylalanine, threonine and tyrosine.These findings are consistent with the strong negative correlation between tannin content and digestibility recorded by Gous et al. (1982).There was no difference (P >0.05) in gross energy content (Table 8) between varieties.
The ME content differed, with the tannin varieties having lower AME and TME contents than the nontannin ones.Amino acid digestibilities, AME and TME of the tannin varieties were low in the present study.Condensed tannins have been reported to reduce feed intake (Oduhu & Baker, 2005;Bryden et al., 2009), metabolizable energy (Bryden et al., 2009;Perez-Maldonado & Rodrigues, 2009;Sannamani et al., 2010;Sedghi et al., 2011) and amino acid digestibility (Selle et al., 2010;Ebadi et al., 2011) in broilers.Such limiting factors are the result of the poor nutritional quality of the protein (kafirins and glutelin) and subsequently attributed to low solubility and deficiencies in amino acids (Sastry et al., 1986;Selle et al., 2010).Therefore, productivity is likely to be reduced when broiler chicks are fed diets containing high concentrations of over 80% of sorghum to replace maize (Oduhu & Baker, 2005;Donald et al., 2008), although tannin-free sorghum might reduce productivity because of factors other than tannin content.

Conclusion
The sorghum varieties NS5511 and PAN8625 contained condensed tannins, while PAN8816 and PAN8906 are non-tannin varieties.There were variations in the nutritional profiles of the four sorghum varieties evaluated in the current study, which could be attributed to factors such as genotype, environment and growing conditions.The tannin sorghum varieties had low amino acid digestibility and metabolizable energy compared with the non-tannin varieties, which suggests the anti-nutritional activity of the tannins.On the other hand, the tannin varieties had higher levels of antioxidants than the non-tannin varieties, thus highlighting their health-promoting potential.It can therefore be concluded that nutrient content and nutritive value should both be considered when selecting sorghum varieties for poultry feeding.
,c,d Means in the same row with different superscripts are significantly different (P <0.05).SEM: standard error of the mean.

Table 2
Phenolic concentration and antioxidant activity of grain of four sorghum varieties

Table 3
Pearson correlation coefficient (r) among kernel characteristics, phenolic content and antioxidant activity of four sorghum varieties

Table 4
Proximate composition (g/kg) of grains of four sorghum varieties (DM basis)

Table 5
Mineral concentrations of grains of four sorghum varieties (mg/kg DM)

Table 6
Amino acid composition of four sorghum varieties (% DM) Can AA be hgher than CP? Means in the same row with different superscripts are significantly different (P <0.05).

Table 7
Amino acid digestibility of grains of four sorghum varieties (% DM) fed to fasted roosters

Table 8
Utilization of energy of four sorghum varieties by roosters Means in the same column with different superscripts are significantly different (P <0.05).