Files
Abstract
Feed accounts for around 70% of poultry production costs, therefore, there is a need to accurately define nutritional values of poultry feed ingredients in order to minimize nutrient deficiencies and/or excesses. One of the most important nutrient specifications to have in poultry diets is the nitrogen corrected true metabolizable energy (TMEN). However, the TMEN animal bioassay is expensive and receiving results can take up to two weeks. Near Infrared Reflectance Spectroscopy (NIRS) has the potential to be used as a tool to predict the TMEN of poultry feed ingredients as it is a rapid analysis method that enables a Multi-Purpose Analyzer to be calibrated to predict nutrient concentrations in a feed ingredient sample. However, the accuracy of a NIRS calibration curve is heavily dependent on the accuracy the associated reference values determined in a laboratory for the NIRS calibration and validation curves. The TMEN bioassay was developed over 30 years ago, and many currently utilized poultry feed ingredients were not investigated, and thus the procedure was not validated for them. The current research aims to validate the rooster TMEN assay and then to use these validated values to construct NIRS calibration and validation curves to predict the TMEN in poultry feed ingredients. For all ingredients tested, a positive correlation (P < 0.0001) existed between N balance and the TME minus TMEN value. The N correction increased the TMEN value relative to its TME value in 10% of the samples and for some of these samples, the N correction inflated the TMEN value by over 20%. Subsequently, an all-encompassing feed ingredient calibration curve was then developed with 150 samples. The all-encompassing validation curve with 551 samples was able to predict with accuracy, with, 96.7% of the predicted TMEN value were within plus or minus 10% of their determined TMEN values. These results suggest the N correction established by the original research is appropriate, but for samples with a negative N balance that exceeds the endogenous control, an adjustment is necessary. These results also indicate that NIRS can predict the TMEN of poultry feed ingredients accurately.