Submitted to: Poultry Science
Publication Type: Peer reviewed journal
Publication Acceptance Date: 4/3/2012
Publication Date: 7/1/2012
Citation: Garcia, R.A., Phillips, J.G., Adeola, O. 2012. Improved prediction of meat and bone meal metabolizable energy content for ducks through in vitro methods. Poultry Science. 91:1854-1859. Interpretive Summary: Meat and bone meal (MBM) is a nutritious substance that is often used to supplement poultry diets. Lack of information on the available energy in MBM, however, makes it difficult for nutritionists to include MBM at the optimal level in the diet. Including more energy than the animal can use is wasteful, while including too little makes the animal grow more slowly. In the present study we developed a method for estimating the amount of available energy in MBM for ducks using simple and inexpensive laboratory methods. Although such efforts have been made in the past, they have had limited success. Our innovation was to include measurements of MBM particle size in the equations used to predict the available energy. Inclusion of these measurements greatly improved these predictions. Such accurate predictions of the available energy in MBM should allow nutritionist to use this feedstuff more easily. Although these results are specific for duck nutrition, we expect that further studies will be able to extend this approach to other poultry species and possibly other types of animals as well.
Technical Abstract: Apparent metabolizable energy (AME) of meat and bone meal (MBM) for poultry is highly variable, but impractical to measure routinely. Previous efforts at developing an in vitro method for predicting AME have had limited success. The present study uses data from a previous publication on the AME of 12 MBM samples, determined using 288 white Peking ducks, as well as composition data on these samples. Here, we investigate the hypothesis that two non-compositional attributes of MBM, particle size and protease resistance, will have utility in improving predictions of AME based on in vitro measurements. Using the same MBM samples as the previous study, protease resistance was determined using a modified pepsin digestibility assay. The samples were also subjected to two measurements of particle size. Analysis of the results using a stepwise construction of multiple linear regression models revealed that the measurements of size were useful predictors of AME, but the measure of protease resistance was not. Relatively simple (4-term) and complex (7-term) models for both AME and nitrogen corrected AME were constructed, with R-squared values ranging from 95.9 to 99.6. The rather minor analytical effort required to conduct the measurements involved is discussed. Although the generality of the results are limited by the number of samples involved and the species used, they suggest that AME for poultry can be accurately predicted through simple and inexpensive in vitro methods.