- Email: [email protected]
Prediction of the true digestible amino acid contents from the chemical composition of sorghum grain for poultry
CALCIUM AND PHOSPHORUS ELOW OE HENS tension. Accessed Mar. 2011. http://edis.ifas.ufl.edu/pdffiles/ VM/VM04400.,pdf..,_ Canadian Agri-Food Research Council. 2003. Recommended eode of practice for the care and handling of pullets, layers, and spent fowl. CARC, Ottawa, Ontario, Canada. Canadian Council on Animal Care. 1993. Cuide to the Care and Use of Experimental Animals. 2nd ed. E. D. Olfert, B. M. Cross, A. A. MeWilliam, ed. CCAC, Ottawa, Ontario, Canada. Chandramoui, B., Jadhao, and R. P. Siuha. 1998. Effect of dietary calcium and phosphorus concentrations on retention of these nutrients by caged layers. Br. Poult. Sei. 39:544-548. De Vries, S., R. P. Kwakkel, and J. Dijkstra. 2010. Dynamics of calcium and phosphorus metabolism in laying hens. Pages 133150 in Phosphorus and Calcium Utilization and Requirements in Farm Animals. D. M. S. S. Vitti and E. Kebreab, ed. CAB International, Wallingford, UK. Duncan, E. T., M. C. Appleby, and B. O. Hughes. 1992. Effect of perches in laying eages on welfare and produetion of hens. Br. Poult. Sei. 33:25-35. Etches, R. J. 1987. Calcium logistics in the laying hen. J. Nutr. 117:619-628. Fleming, R. H., C. C. Whitehead, D. Alvey, N. C. Gregory, and L. J. Wilkins. 1994. Bone structure and breaking strength in laying hens housed in different husbandry systems. Br. Poult. Sei 35:651-662. Gilbert, A. B. 1983. Calcium and reproductive function in the hen. Proc. Nutr. Soc. 42:195-212. Gordon, R. W., and D. A. Roland Sr.. 1998. Influence of supplemental phytase on calcium and phosphorus utilization in laying hens. Poult. Sei. 77:290-294. Guesdon, V., A. M. H. Ahmed, S. Mallet, J. M. Faure, and Y. Nys. 2006. Effects of beak trimming and cage design on laying hen performance and egg quality. Br. Poult. Sei. 47:1-12. Guinotte, F., Y. Nys, and F. De Monredon. 1991. The effects of particle size and origin of calcium carbonate on performance and ossifleation characteristics in broiler chicks. Poult. Sei. 70-19081920. Guru, V. K., S. K. Saxena, and S. P. Netke. 1974. Some effects of confinement on laying pullets. Br. Poult. Sei. 15:333-337. Hester, P. Y. 2005. Impact of science and management on the welfare of egg laying strains of hens. Poult. Sei. 84:687-696. Holder, D. P., and M. V. Bradford. 1979. Relationship of specific gravity of chieken eggs to numbei' of cracked eggs observed and percent shell. Poult. Sei. 58:250-251. Holt, P. S., R. H. Davies, J. Dewulf, R. K. Gast, J. K. Huwe, D. R. Jones, D. Waltman, and K. R. Willian. 2011. The impact of different housing systems on egg safety and quality. Poult. Sei. 90:251-262. Hossain, S. M., and A. G. Bertechini. 1998. Effect of varying manganese and available phosphorus levels in the diet on egg production and eggshell quality of layers. Anim. Feed Sei. Technol. 71:303-308. Hughes, B. O., A. B. Gilbert, and M. F. Brown. 1986. Categorisation and causes of abnormal eggshells: Relationship with stress. Br. Poult. Sei. 27:325-337. Hurwitz, S., and A. Bar. 1967. Calcium metabolism of hens secreting heavy or light eggshells. Poult. Sei. 46:1522-1527. ISA. 2009. Shaver. Nutrition management guide commercials, 2009-10. Accessed Oct. 2010. http://www.isapoultry.com/downloads/l/uutrition_management_guide_commercials_isa_white. pdf. Jendral, M. J., D. R. Korver, J. S. Church, and J. J. R. Feddes. 2008. Bone mineral density and breaking strength of White Leghorns housed in conventional, modifled, and commercially available colony battery cages. Poult. Sei. 87:828-837. Johnston, S. A., and R. M. Gous. 2007. Modelling the changes in the proportions of the egg components during a laying cycle. Br. Poult. Sei. 48:347-353. Jones, D. R., and M. T. Musgrove. 2005. Effects of extended storage on egg quahty factors. Poult. Sei. 84:1774-1777. Kebreab, E., J. France, R. P. Kwakkel, S. Leeson, H. D. Kuhi, and J. Dijkstra. 2009. Development and evaluation of a dynamic model of calcium and phosphorus flows in layers. Poult. Sei. 88:680-689.
2395
Kopka, M. N., H. W. Cheng, and P. Y. Hester. 2003. Bone mineral density of laying hens housed in enriched versus conventional cages. Poult. Sei. 82(Suppl. 1):29. (Abstr.) Lay, D. C. Jr., R. M. Fulton, P. Y. Hester, D. M. Karcher, J. B. Kjaer, J. A. Mench, B. A. Mullens, R. C. Newberry, C. J. Nicol, N. P. O'SuUivan, and R. E. Porter. 2011. Hen welfare in different housing systems. Poult. Sei. 90:278-294. Mench, J. A., A. V. Tienhoven, J. A. Marsh, C. C. McCormic, D. L. Cunningham, and R. C. Baker. 1986. Effects of cage and floor pen management on behavior, production, and physiological stress responses of laying hens. Poult. Sei. 65:1058-1069. Miles, R. D., A. P. T. Costa, and R. H. Harms. 1983. The influence of dietary phosphorus levels on laying hen performance, eggshell quality, and various blood parameters. Poult. Sei. 62:1033-1037. Mueller, W. J., R. Schraer, and H. Schraer. 1964. Calcium metabolism and skeletal dynamics of laying pullets. J. Nutr. 84:20-26. NRC. 1994. Nutrient Requirements of Poultry. 9th ed. National Academy of Sciences, Washington, DC. Neijat, M., J. D. House, W. Cuenter, and E. Kebreab. 2011. Production performance and nitrogen flow of Shaver White layers housed in enriched or conventional caged systems. Poult Sei 90:543-554. N0rgaard-Nielsen, C. 1990. Bone strength of laying hens kept in an alternative system, compared with hens in cages and on deeplitter. Br. Poult. Sei. 31:81-89. Pandey, N. K., C. M. Mahapatra, S. S. Verma, and D. C. Johari. 1986. Effect of strain on physical egg quality characteristics in White Leghorn chickens. Int. J. Poult. Sei. 21:304-307. Parkinson, G., and P. Cransberg. 1999. Early egg production: The relationship between calcium nutrition, appetite, growth, production, and skeletal development. Pages 1-24. A report for the Rural Industries Research and Development Corporation. June 1999. RIRDC Project No: DAV-143A. RIRDC, Attwood, Victoria, Australia. Pelicia, K., E. A. Garcia, A. B. G. Faitarone, A. P. Silva, D. A. Berto, A. B. Molino, and F. Vercese. 2009. Calcium and available phosphorus levels for laying hens in second produetion cycle. Braz. J. Poult. Sei. 11:39-49. Rahn, A. P. 2001. Caged laying hen well-being: An economic perspective. Pages 19-32 in Proc. 52nd North Central Avian Disease Conf. and Symp. The Science Behind Poultry Husbandry, Grand Rapids, MI. Animal Health Diagnostic Laboratory and College of Veterinary Medicine, Michigan State University, Grand Rapids. Rennie, J. S., R. H. Fleming, H. A. McCormack, C. C. McCorquodale, and C. C. Whitehead. 1997. Studies on effects of nutritional factors on bone structure and osteoporosis in laying hens. Br. Poult. Sei. 38:417-424. Reynard, M., and C. J. Savory. 1999. Stress-induced oviposition delays in laying hens: Duration and consequences for eggshell quality. Br. Poult. Sei. 40:585-591. Riezu, C. M., J. L. Saunders-Blades, A. K. Yngvesson, F. E. Robinson, and D. R. Korver. 2004. End-of-cycle bone quality in whiteand brown-egg laying hens. Poult. Sei. 83:375-383. Rodriguez-Navarro, A., O. Kalim, Y. Nys, and J. M. Garcia-Ruiz. 2002. Influence of the microstructure on the shell strength of eggs laid by hens of different ages. Br. Poult. Sei. 43:395-403. Roland, D. A., Sr. 1986. Eggshell quality HI: Calcium and phosphorus requirements of commercial Leghorns. World's Poult. Sei. J. 42:154-163. Roland, D. A., Sr., and R. H. Harms. 1973. Calcium metabohsm in the laying hen. 5. Effect of various sources and size of calcium carbonate on shell quality. Poult. Sei. 55:369-372. Scheideler, S. E., and J. L. Sell. 1986. Effects of calcium and phasefeeding phosphorus on produetion traits and phosphorus retention in two strains of laying hens. Poult. Sei. 65:2110-2119. Scott, T. A., and D. Balnave. 1991. Influence of temperature, dietary energy, nutrient concentration, and self-selection feeding on the retention of dietary energy, protein, and calcium by sexuallymaturing egg-laying pullets. Br. Poult. Sei. 32:1005-1016. Simkiss, K. 1961. Calcium metabolism and avian reproduction. Biol. Rev. Camb. Philos. Soc. 36:321-367. Spiehs, M. J. 2005. Nutritional and feeding strategies to minimize nutrient losses in livestock manure (Ml 188). Accessed Nov. 2010.
2396
NEIJAT ET AL.
http://www.extension.nmn.edu/distribution/livestocksystems/ M1188.html. Summers, J. D., R. Grandhi, and S. Leeson. 1976. Calcium and phosphorus requirements of the laying hen. Poult. Sei. 55:402-413. Taetacan, G. B., W. Guenter, N. J. Lewis, J. G. Rodrignez-Lecompte, and J. D. House. 2009. Performance and welfare of laying hens in conventional and enriched eages. Ponlt. Sei. 88:698-707. Tauson, R. 1998. Health and production in improved cage designs. Poult. Sei. 77:1820-1827. Taylor, T. G. 1965. Dietary phosphorus and eggshell thickness in domestic fowl. Br. Poult. Sei. 6:79-87. Taylor, T. G. 1972. The role of ealeium in egg prodnetion. Pages 107-111 in Egg Formation and Produetion. Symp. 8. Eds. B. M. Freeman, and P. E. Lake. Br. Poult. Sei. Ltd., Edinburgh, Seotland.
'Valkonen, E., E. 'Venalainen, L. Rossow, and J. Valaja. 2010. Effects of ealeium diet supplements on egg strength in eonventional and furnished eages, and effects of 2 different nest floor materials. Ponlt. Sei. 89:2307-2316. 'Vits, A., D. Weitzenburger, H. Hamann, and O. Distl. 2005. Prodnetion, egg quality, bone strength, elaw length, and keel bone deformities of laying hens housed in furnished cages with different group sizes. Poult. Sei. 84:1511-1519. 'Wall, H., R. Tauson, and K. Elwinger. 2002. Effeet of design, passages, and hybrid on use of nest and produetion of layers in furnished eages. Ponlt. Sei. 81:333-339. Whitehead, C. C , and R. H. Fleming. 2000. Osteoporosis in eage layers. Poult. Sei. 79:1033-1041.
Prediction of the true digestible amino acid contents from the chemical composition of sorghum grain for poultry M. R. Ebadi,* M. Sedghi,t^ A. Golian,t and H. Ahmadif department
of Animal Science, Isfahan Research Genter of Agrieulture and Natural Resources, Isfahan, Iran, 81785-199; and fDepartment of Animal Science, Eaculty of Agriculture, Eerdowsi University of Mashhad, Mashhad, Iran, 91775-1163
ABSTRACT Accurate knowledge of true digestible amino acid (TDAA) contents of feedstuffs is necessary to accurately formulate poultry diets for profitable production. Several experimental approaches that are highly expensive and time consuming have been used to determine available amino acids. Prediction of the nutritive value of a feed ingredient from its chemical composition via regression methodology has been attempted for many years. The artificial neural network (ANN) model is a powerful method that may describe the relationship between digestible amino acid contents and chemical composition. Therefore, multiple linear regressions (MLR) and ANN models were developed for predicting the TDAA contents of sorghum grain based on chemical composition. A precision-fed assay trial using cecectomized roosters was performed to determine the TDAA contents in 48 sorghum samples from 12 sorghum varieties differing in chemical composition. The input variables for both MLR and ANN models were CP, ash, crude fiber, ether extract, and total phenols
whereas the output variable was each individual TDAA for every sample. The results of this study revealed that it is possible to satisfactorily estimate the TDAA of sorghum grain through its chemical composition. The chemical composition of sorghum grain seems to highly influence the TDAA contents when considering components such as CP, crude fiber, ether extract, ash and total phenols. It is also possible to estimate the TDAA contents through multiple regression equations with reasonable accuracy depending on composition. However, a more satisfactory prediction may be achieved via ANN for all amino acids. The R-^ values for the ANN model corresponding to testing and training parameters showed a higher accuracy of prediction than equations estabhshed by the MLR method. In addition, the current data confirmed that chemical composition, often considered in total amino acid prediction, could be also a useful predictor of true digestible values of selected amino acids for poultry.
Key words: prediction model, sorghum, true digestible amino acid 2011 Poultry Science 90:2397-2401 doi:10.3382/ps.2011-01413
INTRODUCTION
acid (TDAA) contents of feedstuffs is necessary to formulate poultry diets to achieve profitable production. Sorghum [Sorghum bicolor (L.) Moench] is a drought- In vitro (enzymatic, chemical, and microbiological asresistant grain and thus is an important diet ingredisays), indirect (plasma amino acid assays), and direct ent in semiarid regions of world. Several reports have (digestibility and growth assays) methods have been indicated the adverse effects of some sorghum samples used to determine amino acids availability for poultry on growth rate and feed conversion efficiency of non(Ravindran and Bryden, 1999). Digestibihty trials usruminant animals (Duodu et al., 2003; Ebadi et al., ing live animals have become the most common tech2005). Low protein and amino acid digestibilities of nique for estimating amino acid digestibility but are sorghum grain are the important factors affecting poulexpensive and time consuming. Therefore, nutritiontry performance and production. Because protein and ists are highly interested in finding rapid, inexpensive, amino acids are the most expensive parts of a pouland accurate methods for assessing TDAA contents of try ration, accurate knowledge of true digestible amino feedstuffs. Several studies have shown that amino acid digestibility of sorghum is correlated with its chemical composition (Ebadi et al., 2005; Selle et al., 2010). ©2011 Poultry Science Association Inc. Prediction of nutritive value of a feed ingredient from Received February 7, 2011. its chemical composition has been attempted for many Accepted July 3, 2011. ^Corresponding author: [email protected] years based on regression methods (NRC, 1994; Urriola
2397
2398
EBADI ET AL.
et al., 2009). The interdependencies of the involved factors in such problems limit the use of simple regression analysis and need a more accurate and satisfactory method. The artificial neural network (ANN) model may accurately show the nonlinear behavior with complex relationships and it may estimate the TDAA content in feedstuffs. The ANN model has been applied successfully to the prediction and control of nonlinear systems and systems with unknown models and has been used in poultry feed evaluation (Roush and Cravener, 1997; Cravener and Roush, 1999, 2001; Ahmadi et al., 2008; Perai et al., 2010). Previously, multiple linear regression (MLR) and ANN models have been used to predict the total amino acid content in feed ingredients based on chemical composition (Roush and Cravener, 1997). Information about total amino acid content of feedstuffs is important; however, it is more essential for a nutritionist to know the TDAA contents for specific feed ingredients when formulating poultry diets. Differences in chemical composition may be responsible for a substantial amount of the variability in the nutritional value of sorghum grain that can affect amino acid content and digestibility. Because of a high level of antinutritional factors, amino acid digestibility is a main concern in the use of sorghum as a feed component for poultry. The relationship between sorghum chemical composition and TDAA content is not yet documented. Therefore, the main objective of this study was to develop MLR and ANN models for poultry to predict the sorghum grain TDAA content based on its chemical composition.
MATERIALS AND METHODS Data Collection Twelve varieties of sorghum grain were grown under similar environmental conditions. Four samples from each of the 12 varieties were taken and evaluated for CP, ash, crude fiber (CF), and ether extract (EE) following AOAC International (2000) analytical methods (990.03, 920.39, 978.10, and 942.05, respectively). Total phenolics were assayed by the Folin-Denis method (952.03; AOAC International, 2000). Amino acid concentrations in sorghum grain samples were analyzed by ion exchange chromatography following hydrolysis in 6 A^ HCl for 24 h at 110°C in sealed tubes, with at least 2 replicates per sample. Derivation with ninhydrin was accomplished (Andrews and Baldar, 1985) and the quantity of each amino acid was determined using the Bechman Biochrom 20 Amino Acid Analyzer at the University of Manitoba (Winnipeg, Canada). Methionine and Cys were determined on samples that had been oxidized in performic acid before acid hydrolysis (Moore, 1963). This project was approved by the Animal Care Committee of the Ferdowsi University of Mashhad, Iran. Single comb White Leghorn roosters were cecectomized according to Parson's method (Parsons, 1985). After a
recovery period and 24 h of feed deprivation, roosters were randomly given a 30-g sorghum sample via crop intubation (6 roosters/sorghum sample). Six additional roosters were fed 30 g of glucose to measure endogenous amino acids excreted (Creen et al., 1987; McNab and Blair, 1988). The excreta were collected over a 48-h period and stored in a freezer. After all excreta were freeze dried, the concentrations of amino acids were analyzed as described previously. True amino acid digestibility coefficient was calculated by the Sibbald (1986) method. The TDAA values were obtained by multiplying total amino acid contents and true amino acid digestibility coefficients. In the present study, 16 amino acids (Asp, Thr, Ser, Glu, Pro, Ala, Cys, Val, Met, He, Leu, Tyr, Phe, His, Lys, and Arg) were selected to assess the relationship between these amino acid digestibilities and the sorghum chemical compositions.
Model Development Two methods of MLR and ANN models were used to predict TDAA contents in 48 sorghum grain samples. The input variables for both MLR and ANN models were CP, CF, EE, ash, and total phenols. Each individual TDAA content was the output variable. To avoid any bias, the 48 data lines were randomly split into training and testing sets with 34 and 14 data lines, respectively. An algorithm of feed-forward multilayer perceptron with 5 inputs, 1 output (with a linear activation function), and 6 hidden neurons (with a hyperbolic tangent activation function) was considered to construct the ANN model. A training algorithm of Quasi-Newton was used to train the network (Lou and Nakai, 2001; Ahmadi and Gohan, 2010). The ANN models were made using Statistica Neural Networks software (version 8.0; StatSoft, 2009). Evaluation of model performance was based on the accuracy of the prediction on the testing data. Data from the training set (34 data lines) were also fitted for linear regression using SAS PROC REC (SAS Institute, 2003). Quantitative examination of the predictive ability of both MLR and ANN models was determined using R^, MS error, and bias for each amino acid (Roush et al., 2006).
RESULTS AND DISCUSSION Average, minimum, maximum, and SD values (n = 48) of chemical composition and TDAA content for all 48 sorghum samples are shown in Table 1. The results indicate a great variability in both the sorghum chemical composition and TDAA content obtained in sorghum grain. The difference between maximum and minimum values for TDAA content of sorghum samples suggests that the formulation of diets based on average nutritive value may not be the most accurate method. The MLR and ANN models prediction efficiency, as R^, MS error, and bias obtained for each amino acid and chosen ANN models architecture, is shown
PREDICTION MODEL FOR DIGESTIBLE AMINO ACIDS IN SORGHUM CD Ci Oi CO OI I-H 'Tf o CO i-H íN
4-3
Ö oÖÖ
a;
2399 O oo Ci q q
Ö Ö cS CO es rH es o lo q q
O OOÖ
S ö ö
O O 00 o "-H Ci q q ó d Ö
o es ^ oo 00 q q Ö d> Ö I
S
o oo o
1—1 o iO o o b~ o o
'3
es o 00 rH o t^ o o
o öö ö
oo rH o o oi o o
o Ö Ö
O CO ^ CO es Cl o LO
CD es
"^oo ooqq có d) Ö
IO o C O I-H
o o ö ö
CO 00
^ o o oi o o
Ö o Ö
O O
Lo-^es Cioo d d d
Oioo oiOo d dd
!-< es LO CO o d dd
oo Oi o o d dd
O> o rt 05 o o Ö ÖÖ
i-H (ÜD 00 T-H
ÍD (M rH tM 00 CO rH lO rH Ö 1-H Ö
§
l>- CO Oi CN Oi es b- CO lO es CO 1—1
ö öö ö 1-H to es d>
"^ IO q dl
Ci -^ I—I Ö
es "^ ) LO o
00 LO o Ö
fa o
o ö dö -3
Ci b- CD I-H T-H C)0 "^ l O es o r-H q
00 00 q o Ö Ö Ö
a '3
c: o o Ö Ö Ö
d d d
d dd
ce o Oi o -H Oí o o
lo 10 CD t-- ,—I 00 q q
00 t^ o o Oi o o
01 o o
d d d
dd d
^
OJ
o: q q Ö ci Ö
S
d öö ö O
Oî o o
o; q q d dd
t— Oi Ci CO 00 00 ÍD -^ ^ -^ Ci CO
c
d d Ö
03 00 00 i-H 00 ^ b- t- '^ es '-H CO CO
3
1-3 d d d
00 q q Ö Ö Ö
t 00 o 00 q q
d d d
01 -O
o
I>- o CO 00 Ci es "^ o
o
i-H CD C5 00
o o
o to
CO o o Oî q q
§88
ö ó Ö Oi i-H
O
Ci '^
C^ CO
I—I T^ o 0 0 0 0
"^ CO CO Cieso
00 o o c i o o
CO o lO r-~i o ooqq 000
CO "^ • ^ CO I>Cii—iq d d d
es es 00 o o CiOO d d d
^00 to 00 es d d d
CD o eTi o o d dd
"^00 incoes
rH o t-00
á
d> <ó I
CN CO O5 O es Oí O O
tí
13
o oo o CL, 5 CD 00 -^ J CO CD es
C 00 o 00 q q Ö Ö Ö
Q M
CD 00
> O es
bO 0)
bO
"^ o 0 0 0
Ö
Oi 00 es Oi
j
; p^ CO 00 00 CO
o 00 ^ es
es es es CO
Ci o -^ d
lO es 00 CO
00 Ci 00 d
lO 00 Ci es
-^ Ci es rH
II I
) o o
coqq 000
en
1
Ciqq ood I
o o
CN o rH o o 00 q q Ö Ö ¡ó IV o to q q
ó ÖÖ
Ciqo d dd
rH ^
Oí o es
o; q q à Ö <ó rH CO
CO O es Oí q q
Ö óÖ o o
•-^ o
01 o o Oí o o
•a , bO
a
9 < -1-3
d, wi
03 o CN OÍ o o
Ci Ol Oi I CO f-H rH
q CO e^ Ö ÖÖ Ö
s¿ TU
b
3S S <; Oí
¡¿Í;
II
Copyright of Poultry Science is the property of Poultry Science Association, Inc. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
2395
Kopka, M. N., H. W. Cheng, and P. Y. Hester. 2003. Bone mineral density of laying hens housed in enriched versus conventional cages. Poult. Sei. 82(Suppl. 1):29. (Abstr.) Lay, D. C. Jr., R. M. Fulton, P. Y. Hester, D. M. Karcher, J. B. Kjaer, J. A. Mench, B. A. Mullens, R. C. Newberry, C. J. Nicol, N. P. O'SuUivan, and R. E. Porter. 2011. Hen welfare in different housing systems. Poult. Sei. 90:278-294. Mench, J. A., A. V. Tienhoven, J. A. Marsh, C. C. McCormic, D. L. Cunningham, and R. C. Baker. 1986. Effects of cage and floor pen management on behavior, production, and physiological stress responses of laying hens. Poult. Sei. 65:1058-1069. Miles, R. D., A. P. T. Costa, and R. H. Harms. 1983. The influence of dietary phosphorus levels on laying hen performance, eggshell quality, and various blood parameters. Poult. Sei. 62:1033-1037. Mueller, W. J., R. Schraer, and H. Schraer. 1964. Calcium metabolism and skeletal dynamics of laying pullets. J. Nutr. 84:20-26. NRC. 1994. Nutrient Requirements of Poultry. 9th ed. National Academy of Sciences, Washington, DC. Neijat, M., J. D. House, W. Cuenter, and E. Kebreab. 2011. Production performance and nitrogen flow of Shaver White layers housed in enriched or conventional caged systems. Poult Sei 90:543-554. N0rgaard-Nielsen, C. 1990. Bone strength of laying hens kept in an alternative system, compared with hens in cages and on deeplitter. Br. Poult. Sei. 31:81-89. Pandey, N. K., C. M. Mahapatra, S. S. Verma, and D. C. Johari. 1986. Effect of strain on physical egg quality characteristics in White Leghorn chickens. Int. J. Poult. Sei. 21:304-307. Parkinson, G., and P. Cransberg. 1999. Early egg production: The relationship between calcium nutrition, appetite, growth, production, and skeletal development. Pages 1-24. A report for the Rural Industries Research and Development Corporation. June 1999. RIRDC Project No: DAV-143A. RIRDC, Attwood, Victoria, Australia. Pelicia, K., E. A. Garcia, A. B. G. Faitarone, A. P. Silva, D. A. Berto, A. B. Molino, and F. Vercese. 2009. Calcium and available phosphorus levels for laying hens in second produetion cycle. Braz. J. Poult. Sei. 11:39-49. Rahn, A. P. 2001. Caged laying hen well-being: An economic perspective. Pages 19-32 in Proc. 52nd North Central Avian Disease Conf. and Symp. The Science Behind Poultry Husbandry, Grand Rapids, MI. Animal Health Diagnostic Laboratory and College of Veterinary Medicine, Michigan State University, Grand Rapids. Rennie, J. S., R. H. Fleming, H. A. McCormack, C. C. McCorquodale, and C. C. Whitehead. 1997. Studies on effects of nutritional factors on bone structure and osteoporosis in laying hens. Br. Poult. Sei. 38:417-424. Reynard, M., and C. J. Savory. 1999. Stress-induced oviposition delays in laying hens: Duration and consequences for eggshell quality. Br. Poult. Sei. 40:585-591. Riezu, C. M., J. L. Saunders-Blades, A. K. Yngvesson, F. E. Robinson, and D. R. Korver. 2004. End-of-cycle bone quality in whiteand brown-egg laying hens. Poult. Sei. 83:375-383. Rodriguez-Navarro, A., O. Kalim, Y. Nys, and J. M. Garcia-Ruiz. 2002. Influence of the microstructure on the shell strength of eggs laid by hens of different ages. Br. Poult. Sei. 43:395-403. Roland, D. A., Sr. 1986. Eggshell quality HI: Calcium and phosphorus requirements of commercial Leghorns. World's Poult. Sei. J. 42:154-163. Roland, D. A., Sr., and R. H. Harms. 1973. Calcium metabohsm in the laying hen. 5. Effect of various sources and size of calcium carbonate on shell quality. Poult. Sei. 55:369-372. Scheideler, S. E., and J. L. Sell. 1986. Effects of calcium and phasefeeding phosphorus on produetion traits and phosphorus retention in two strains of laying hens. Poult. Sei. 65:2110-2119. Scott, T. A., and D. Balnave. 1991. Influence of temperature, dietary energy, nutrient concentration, and self-selection feeding on the retention of dietary energy, protein, and calcium by sexuallymaturing egg-laying pullets. Br. Poult. Sei. 32:1005-1016. Simkiss, K. 1961. Calcium metabolism and avian reproduction. Biol. Rev. Camb. Philos. Soc. 36:321-367. Spiehs, M. J. 2005. Nutritional and feeding strategies to minimize nutrient losses in livestock manure (Ml 188). Accessed Nov. 2010.
2396
NEIJAT ET AL.
http://www.extension.nmn.edu/distribution/livestocksystems/ M1188.html. Summers, J. D., R. Grandhi, and S. Leeson. 1976. Calcium and phosphorus requirements of the laying hen. Poult. Sei. 55:402-413. Taetacan, G. B., W. Guenter, N. J. Lewis, J. G. Rodrignez-Lecompte, and J. D. House. 2009. Performance and welfare of laying hens in conventional and enriched eages. Ponlt. Sei. 88:698-707. Tauson, R. 1998. Health and production in improved cage designs. Poult. Sei. 77:1820-1827. Taylor, T. G. 1965. Dietary phosphorus and eggshell thickness in domestic fowl. Br. Poult. Sei. 6:79-87. Taylor, T. G. 1972. The role of ealeium in egg prodnetion. Pages 107-111 in Egg Formation and Produetion. Symp. 8. Eds. B. M. Freeman, and P. E. Lake. Br. Poult. Sei. Ltd., Edinburgh, Seotland.
'Valkonen, E., E. 'Venalainen, L. Rossow, and J. Valaja. 2010. Effects of ealeium diet supplements on egg strength in eonventional and furnished eages, and effects of 2 different nest floor materials. Ponlt. Sei. 89:2307-2316. 'Vits, A., D. Weitzenburger, H. Hamann, and O. Distl. 2005. Prodnetion, egg quality, bone strength, elaw length, and keel bone deformities of laying hens housed in furnished cages with different group sizes. Poult. Sei. 84:1511-1519. 'Wall, H., R. Tauson, and K. Elwinger. 2002. Effeet of design, passages, and hybrid on use of nest and produetion of layers in furnished eages. Ponlt. Sei. 81:333-339. Whitehead, C. C , and R. H. Fleming. 2000. Osteoporosis in eage layers. Poult. Sei. 79:1033-1041.
Prediction of the true digestible amino acid contents from the chemical composition of sorghum grain for poultry M. R. Ebadi,* M. Sedghi,t^ A. Golian,t and H. Ahmadif department
of Animal Science, Isfahan Research Genter of Agrieulture and Natural Resources, Isfahan, Iran, 81785-199; and fDepartment of Animal Science, Eaculty of Agriculture, Eerdowsi University of Mashhad, Mashhad, Iran, 91775-1163
ABSTRACT Accurate knowledge of true digestible amino acid (TDAA) contents of feedstuffs is necessary to accurately formulate poultry diets for profitable production. Several experimental approaches that are highly expensive and time consuming have been used to determine available amino acids. Prediction of the nutritive value of a feed ingredient from its chemical composition via regression methodology has been attempted for many years. The artificial neural network (ANN) model is a powerful method that may describe the relationship between digestible amino acid contents and chemical composition. Therefore, multiple linear regressions (MLR) and ANN models were developed for predicting the TDAA contents of sorghum grain based on chemical composition. A precision-fed assay trial using cecectomized roosters was performed to determine the TDAA contents in 48 sorghum samples from 12 sorghum varieties differing in chemical composition. The input variables for both MLR and ANN models were CP, ash, crude fiber, ether extract, and total phenols
whereas the output variable was each individual TDAA for every sample. The results of this study revealed that it is possible to satisfactorily estimate the TDAA of sorghum grain through its chemical composition. The chemical composition of sorghum grain seems to highly influence the TDAA contents when considering components such as CP, crude fiber, ether extract, ash and total phenols. It is also possible to estimate the TDAA contents through multiple regression equations with reasonable accuracy depending on composition. However, a more satisfactory prediction may be achieved via ANN for all amino acids. The R-^ values for the ANN model corresponding to testing and training parameters showed a higher accuracy of prediction than equations estabhshed by the MLR method. In addition, the current data confirmed that chemical composition, often considered in total amino acid prediction, could be also a useful predictor of true digestible values of selected amino acids for poultry.
Key words: prediction model, sorghum, true digestible amino acid 2011 Poultry Science 90:2397-2401 doi:10.3382/ps.2011-01413
INTRODUCTION
acid (TDAA) contents of feedstuffs is necessary to formulate poultry diets to achieve profitable production. Sorghum [Sorghum bicolor (L.) Moench] is a drought- In vitro (enzymatic, chemical, and microbiological asresistant grain and thus is an important diet ingredisays), indirect (plasma amino acid assays), and direct ent in semiarid regions of world. Several reports have (digestibility and growth assays) methods have been indicated the adverse effects of some sorghum samples used to determine amino acids availability for poultry on growth rate and feed conversion efficiency of non(Ravindran and Bryden, 1999). Digestibihty trials usruminant animals (Duodu et al., 2003; Ebadi et al., ing live animals have become the most common tech2005). Low protein and amino acid digestibilities of nique for estimating amino acid digestibility but are sorghum grain are the important factors affecting poulexpensive and time consuming. Therefore, nutritiontry performance and production. Because protein and ists are highly interested in finding rapid, inexpensive, amino acids are the most expensive parts of a pouland accurate methods for assessing TDAA contents of try ration, accurate knowledge of true digestible amino feedstuffs. Several studies have shown that amino acid digestibility of sorghum is correlated with its chemical composition (Ebadi et al., 2005; Selle et al., 2010). ©2011 Poultry Science Association Inc. Prediction of nutritive value of a feed ingredient from Received February 7, 2011. its chemical composition has been attempted for many Accepted July 3, 2011. ^Corresponding author: [email protected] years based on regression methods (NRC, 1994; Urriola
2397
2398
EBADI ET AL.
et al., 2009). The interdependencies of the involved factors in such problems limit the use of simple regression analysis and need a more accurate and satisfactory method. The artificial neural network (ANN) model may accurately show the nonlinear behavior with complex relationships and it may estimate the TDAA content in feedstuffs. The ANN model has been applied successfully to the prediction and control of nonlinear systems and systems with unknown models and has been used in poultry feed evaluation (Roush and Cravener, 1997; Cravener and Roush, 1999, 2001; Ahmadi et al., 2008; Perai et al., 2010). Previously, multiple linear regression (MLR) and ANN models have been used to predict the total amino acid content in feed ingredients based on chemical composition (Roush and Cravener, 1997). Information about total amino acid content of feedstuffs is important; however, it is more essential for a nutritionist to know the TDAA contents for specific feed ingredients when formulating poultry diets. Differences in chemical composition may be responsible for a substantial amount of the variability in the nutritional value of sorghum grain that can affect amino acid content and digestibility. Because of a high level of antinutritional factors, amino acid digestibility is a main concern in the use of sorghum as a feed component for poultry. The relationship between sorghum chemical composition and TDAA content is not yet documented. Therefore, the main objective of this study was to develop MLR and ANN models for poultry to predict the sorghum grain TDAA content based on its chemical composition.
MATERIALS AND METHODS Data Collection Twelve varieties of sorghum grain were grown under similar environmental conditions. Four samples from each of the 12 varieties were taken and evaluated for CP, ash, crude fiber (CF), and ether extract (EE) following AOAC International (2000) analytical methods (990.03, 920.39, 978.10, and 942.05, respectively). Total phenolics were assayed by the Folin-Denis method (952.03; AOAC International, 2000). Amino acid concentrations in sorghum grain samples were analyzed by ion exchange chromatography following hydrolysis in 6 A^ HCl for 24 h at 110°C in sealed tubes, with at least 2 replicates per sample. Derivation with ninhydrin was accomplished (Andrews and Baldar, 1985) and the quantity of each amino acid was determined using the Bechman Biochrom 20 Amino Acid Analyzer at the University of Manitoba (Winnipeg, Canada). Methionine and Cys were determined on samples that had been oxidized in performic acid before acid hydrolysis (Moore, 1963). This project was approved by the Animal Care Committee of the Ferdowsi University of Mashhad, Iran. Single comb White Leghorn roosters were cecectomized according to Parson's method (Parsons, 1985). After a
recovery period and 24 h of feed deprivation, roosters were randomly given a 30-g sorghum sample via crop intubation (6 roosters/sorghum sample). Six additional roosters were fed 30 g of glucose to measure endogenous amino acids excreted (Creen et al., 1987; McNab and Blair, 1988). The excreta were collected over a 48-h period and stored in a freezer. After all excreta were freeze dried, the concentrations of amino acids were analyzed as described previously. True amino acid digestibility coefficient was calculated by the Sibbald (1986) method. The TDAA values were obtained by multiplying total amino acid contents and true amino acid digestibility coefficients. In the present study, 16 amino acids (Asp, Thr, Ser, Glu, Pro, Ala, Cys, Val, Met, He, Leu, Tyr, Phe, His, Lys, and Arg) were selected to assess the relationship between these amino acid digestibilities and the sorghum chemical compositions.
Model Development Two methods of MLR and ANN models were used to predict TDAA contents in 48 sorghum grain samples. The input variables for both MLR and ANN models were CP, CF, EE, ash, and total phenols. Each individual TDAA content was the output variable. To avoid any bias, the 48 data lines were randomly split into training and testing sets with 34 and 14 data lines, respectively. An algorithm of feed-forward multilayer perceptron with 5 inputs, 1 output (with a linear activation function), and 6 hidden neurons (with a hyperbolic tangent activation function) was considered to construct the ANN model. A training algorithm of Quasi-Newton was used to train the network (Lou and Nakai, 2001; Ahmadi and Gohan, 2010). The ANN models were made using Statistica Neural Networks software (version 8.0; StatSoft, 2009). Evaluation of model performance was based on the accuracy of the prediction on the testing data. Data from the training set (34 data lines) were also fitted for linear regression using SAS PROC REC (SAS Institute, 2003). Quantitative examination of the predictive ability of both MLR and ANN models was determined using R^, MS error, and bias for each amino acid (Roush et al., 2006).
RESULTS AND DISCUSSION Average, minimum, maximum, and SD values (n = 48) of chemical composition and TDAA content for all 48 sorghum samples are shown in Table 1. The results indicate a great variability in both the sorghum chemical composition and TDAA content obtained in sorghum grain. The difference between maximum and minimum values for TDAA content of sorghum samples suggests that the formulation of diets based on average nutritive value may not be the most accurate method. The MLR and ANN models prediction efficiency, as R^, MS error, and bias obtained for each amino acid and chosen ANN models architecture, is shown
PREDICTION MODEL FOR DIGESTIBLE AMINO ACIDS IN SORGHUM CD Ci Oi CO OI I-H 'Tf o CO i-H íN
4-3
Ö oÖÖ
a;
2399 O oo Ci q q
Ö Ö cS CO es rH es o lo q q
O OOÖ
S ö ö
O O 00 o "-H Ci q q ó d Ö
o es ^ oo 00 q q Ö d> Ö I
S
o oo o
1—1 o iO o o b~ o o
'3
es o 00 rH o t^ o o
o öö ö
oo rH o o oi o o
o Ö Ö
O CO ^ CO es Cl o LO
CD es
"^oo ooqq có d) Ö
IO o C O I-H
o o ö ö
CO 00
^ o o oi o o
Ö o Ö
O O
Lo-^es Cioo d d d
Oioo oiOo d dd
!-< es LO CO o d dd
oo Oi o o d dd
O> o rt 05 o o Ö ÖÖ
i-H (ÜD 00 T-H
ÍD (M rH tM 00 CO rH lO rH Ö 1-H Ö
§
l>- CO Oi CN Oi es b- CO lO es CO 1—1
ö öö ö 1-H to es d>
"^ IO q dl
Ci -^ I—I Ö
es "^ ) LO o
00 LO o Ö
fa o
o ö dö -3
Ci b- CD I-H T-H C)0 "^ l O es o r-H q
00 00 q o Ö Ö Ö
a '3
c: o o Ö Ö Ö
d d d
d dd
ce o Oi o -H Oí o o
lo 10 CD t-- ,—I 00 q q
00 t^ o o Oi o o
01 o o
d d d
dd d
^
OJ
o: q q Ö ci Ö
S
d öö ö O
Oî o o
o; q q d dd
t— Oi Ci CO 00 00 ÍD -^ ^ -^ Ci CO
c
d d Ö
03 00 00 i-H 00 ^ b- t- '^ es '-H CO CO
3
1-3 d d d
00 q q Ö Ö Ö
t 00 o 00 q q
d d d
01 -O
o
I>- o CO 00 Ci es "^ o
o
i-H CD C5 00
o o
o to
CO o o Oî q q
§88
ö ó Ö Oi i-H
O
Ci '^
C^ CO
I—I T^ o 0 0 0 0
"^ CO CO Cieso
00 o o c i o o
CO o lO r-~i o ooqq 000
CO "^ • ^ CO I>Cii—iq d d d
es es 00 o o CiOO d d d
^00 to 00 es d d d
CD o eTi o o d dd
"^00 incoes
rH o t-00
á
d> <ó I
CN CO O5 O es Oí O O
tí
13
o oo o CL, 5 CD 00 -^ J CO CD es
C 00 o 00 q q Ö Ö Ö
Q M
CD 00
> O es
bO 0)
bO
"^ o 0 0 0
Ö
Oi 00 es Oi
j
; p^ CO 00 00 CO
o 00 ^ es
es es es CO
Ci o -^ d
lO es 00 CO
00 Ci 00 d
lO 00 Ci es
-^ Ci es rH
II I
) o o
coqq 000
en
1
Ciqq ood I
o o
CN o rH o o 00 q q Ö Ö ¡ó IV o to q q
ó ÖÖ
Ciqo d dd
rH ^
Oí o es
o; q q à Ö <ó rH CO
CO O es Oí q q
Ö óÖ o o
•-^ o
01 o o Oí o o
•a , bO
a
9 < -1-3
d, wi
03 o CN OÍ o o
Ci Ol Oi I CO f-H rH
q CO e^ Ö ÖÖ Ö
s¿ TU
b
3S S <; Oí
¡¿Í;
II
Copyright of Poultry Science is the property of Poultry Science Association, Inc. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.