Explore
Communities in English
Advertise on Engormix

Assessing AME and Digestible Amino Acids of Different Soybean Meals by NIRS and Broiler Performance

Published: November 12, 2012
By: Y. G. Liu (Adisseo Asia Pacific Pte Ltd,Singapore), Bob Swick(University of New England, UNE, Australia) and David Creswell (Creswell Nutrition, Sydney, Australia)
Summary

It has long been realized that the nutritive values of major feed ingredients vary considerably not only among different feedstuff but also the same ingredients of different origins and production batches. Rapid screening to assess true nutritive value has been a great challenge in the feed industry. This study investigated soybean meal (SBM) samples of four different origins using both near infrared reflectance spectroscopy (NIRS) and broiler experiments. In Broiler Test 1, 4 SBM samples were assigned an identical "book values", so that the four dietary formulations were the same except different sources of SBM; In Broiler Test 2, individual NIRS prediction values were used for formulation so SBM inclusion levels varied to ensure diets were iso-energetic and iso-nitrogenous. The broiler studies consisted in eight treatments with 6 replicates using 16 male birds each, from day 1 to 40 of age. The results showed that when using the common values for the four SBM, broiler performed very differently in terms of live weight gain and feed conversion (P<0.05); whilst formulating based on individual NIRS values, the birds grew very similarly. The results clearly demonstrate that the nutritive values of the four SBM varied considerably among origins and largely differed by their levels of metabolisable energy (ME) and digestible amino acids (DAA) rather than crude protein level, and that the NIRS estimates for AME and DAA appeared to be largely correct.

I. INTRODUCTION
Numerous works have shown substantial variation in nutritional value of soybean meals (SBM) due to several factors such as genotype, crushing, anti-nutritional factors, implying a need for better and rapid characterization (De Coca-Sinova et al., 2008; Karr-Lilienthal et al., 2004). During the past two decades, Adisseo has been conducting in vivo studies to evaluate digestibility of amino acids (Green et al., 1987) and energy (Bourdillon et al., 1990) of SBM, the data were used to develop prediction models using the Near Infrared Reflectance Spectroscopy (NIRS), and validated by both in vitro (high performance liquid chromatography or HPLC) and in vivo tests. These NIRS models can be used to predict nutritional value of SBM with a reasonably good level of accuracy (Tang et al., 2008).
The objective of the present study was to assess whether the NIRS is a more practical and accurate approach of nutritional assessment of SBM than average table values on nutrient composition of these ingredients. 
II. MATERIALS AND METHODS
Through a collaboration with America Soybean Association (ASA), we collected four samples of SBM from 4 major producing countries (non-dehulled SBM from India, dehulled SBM from Argentina, Malaysia and USA), and predicted its nutritive values (Table 1) using NIRS equations for proximate nutrients, total and digestible amino acids and apparent metabolisable energy (AME).
Two sets of broiler performance experiments were carried out at Bangkok Animal Research Centre (BARC, Thailand) to evaluate the values of those 4 SBM, using corn-cassava-soybean meals-based diets. Besides SBM, all the other ingredients were identical and at the same inclusion levels in the formulations. The specifications of SBM, on the other hand, were based on either common SBM specifications (USA) or individual NIRS specifications.
The experiment used 768 newly hatched Arbor Acres FSY male broiler chicks allocated to a completely randomised design, each treatment consisted in 6 replicates and 16 male birds per replicate. The experiment was conducted in a closed house with tunnel ventilation and evaporative cooling system. Birds were raised on concrete-floor pens using rice hull as bedding material. Feed and water were provided ad libitum. The diets were provided in crumble form during the first 10 days and pellet form thereafter. Total pen feed consumption was recorded weekly. Body weight and feed intake as pen basis was measured for growth and feed conversion ratio calculation at 18 and 40 days of age. Mortality was recorded daily. Body weight gain, feed intake, feed conversion ratio and livability were calculated and were subjected to analysis of variance in randomized complete block design.
All formulations were iso-energetic and iso-nitrogenous, based on digestible amino acids. The starter diets contained ME 12.74 MJ/kg or 2950 kcal/kg, protein 211 mg/kg and digestible lysine 11.5 mg/kg (Table 2) and the grower diets had ME 12.96 MJ/kg or 3,000 kcal/kg, crude protein 190 mg/kg and digestible lysine 10.0 mg/kg (formulations not shown).
Table 1. Nutrient values (mg/kg) of the four soybean meals as estimated by NIRS
Assessing AME and Digestible Amino Acids of Different Soybean Meals by NIRS and Broiler Performance - Image 1
III. RESULTS AND DISCUSSION
Results presented in Table 3 illustrate a generally good performance. Overall performance in this trial was excellent; with the best treatment achieving live weight of 2908 g and FCR 1.624 at 40 days (Ross 308 standard for males is 2669 g and 1.664 FCR at 40 days), with a precision found at 2% variation for most performance parameters.
When all four soybean meals were formulated with the same soybean meal specification (US values, Treatments 1-4), broilers performed quite differently among treatments (P<0.05), ranking SBM values from low to high in the following order: India, Argentina, USA and Malaysia origins. The Indian SBM showed the poorest quality with 14 points higher in FCR and 100 g lower live weight as compared to the highest quality SBM of Malaysia origin.
On the contrary, when formulating with NIRS specifications (Treatments 5-8), the differences among the 4 SBM samples were minimised, all birds achieved similar live weight and FCR among the four treatments. These results suggest that the NIRS values be closer to the true nutritive values of the SBM than simply using book values since little difference among treatments resulted from formulating using the NIRS values.
Table 2. Composition of starter diets (1-18 days) *
Assessing AME and Digestible Amino Acids of Different Soybean Meals by NIRS and Broiler Performance - Image 2
Table 3. Performance of male broilers (0-40 days) fed on different soybean meals based on common (US) and NIRS specifications
Assessing AME and Digestible Amino Acids of Different Soybean Meals by NIRS and Broiler Performance - Image 3
The economic values of the 4 soybean meals were estimated based on market prices of the feed ingredients and broiler carcasses. Results showed their values were quite different, with as much as U$120/t difference between the top and bottom quality meals, of which over half came from the difference in AME values. 
IV. DISCUSSION
This study result demonstrated that the 4 commercial soybean meals used are very different in their nutritive quality, as shown by NIRS prediction and their performances in the broiler diets. The data also confirmed that the quality of a soybean meal is largely defined by its levels of metabolizable energy (ME) and digestible amino acids, rather than crude protein levels as routinely judged in the most commercial practice.
When soybean meals of different qualities are formulated with their correct specifications of energy and digestible amino acids, they would be expected to produce similar levels of live performance. It means a lower quality soybean meal (with lower levels of energy and digestible amino acids) are able to produce good broiler performance if it is formulated with its correct nutrient values.
The NIRS determined estimates for energy and digestible amino acids as conducted for this trial appear to be largely correct, because when these values were used in formulation, the birds performed similarly.
A rather surprising finding from this trial was the large differences in energy between the soybean meals and the economic value of the energy differences. For example, between Malaysia and Indian meals there was almost 1.674 MJ/kg difference in ME, and the US and Argentine meals had about 0.84 MJ/kg ore ME than the Indian meal. This was picked up by NIRS, and confirmed by the performance results when these soybean meals were fed, indicating that the energy value of a soybean meal can vary greatly and has a large impact on its economic value. 
V. CONCLUSION
In conclusion, the results clearly demonstrated that the nutritive quality of soybean meals is not the same, and their values are largely defined by their levels of metabolisable energy (ME) and digestible amino acids (DAA). This means that if the levels of ME and DAA of any soybean meal are known, then the quality and economic value of that meal can be calculated. A lower quality soybean meal (with lower levels of ME and DAA) may still produce good broiler performance if it is formulated with its correct nutrient values. The NIRS technology for ME and DAA estimation as conducted for this trial is promising. 
REFERENCES
Bourdillon A, Carré B, Conan L, Duperray, JJ, Huyghebaert G, Leclercq B, Lessire M, McNab J , Wiseman J, (1990) Br. Poult. Sci., 31, 557-565.
Creswell D and Swick R (2009) Asian Poultry April, 20-25.
De Coca-Sinova A, Valencia DG, Jiménez-Moreno E, Lázaro R, Mateos GG (2008) Poult. Sci. 87, 2613-2623.
Green S, Bertrand SL, Duron MJC, Maillard R (1987) Br. Poult. Sci. 28, 631-641.
Karr-Lilienthal LK, Merchen NR, Grieshop CM, Flahaven MA, Mahan DC, Fastinger ND, Watts M, Fahey Jr. GC (2004) J Anim Sci. 82, 3198-3209.
Tang L, Liu YG, Gady C, Geraert PA (2008) Soybean Meal Quality Symposium, Bangkok
Related topics
Authors:
Robert Swick
University of New England
Follow
David Creswell
Follow
Join to be able to comment.
Once you join Engormix, you will be able to participate in all content and forums.
* Required information
Would you like to discuss another topic? Create a new post to engage with experts in the community.
Create a post
Nnaemeka Anyanwu
14 de noviembre de 2012

The content of protein for the for the starter and grower diets was given as 211mg/kg and 190mg/kg under materials and methods. But I think that it should be 21% and 19% respectively for a broiler starter and grower phase, in that case it should be 211g/kg and 190g/kg respectively. please correct me if am wrong! Thanks.

Join Engormix and be part of the largest agribusiness social network in the world.
LoginRegister