Evaluation of near-infrared reflectance spectroscopy (NIRS) techniques for total and phytate phosphorus of common poultry feed ingredients.

The purpose of this study was to determine the feasibility of estimating the total and phytate P content of common poultry feed ingredients by near-infrared reflectance spectroscopy (NIRS). Samples of 8 plant-origin feedstuffs were collected from poultry producers in the USA and Canada during the summer of 2009: corn (133), soybean meal (114), corn distillers dried grains with solubles (DDGS; 89), bakery by-product meal (95), wheat (22), wheat middlings (31), canola meal (21), and wheat shorts (15). The samples were assayed by standard wet chemical techniques for total and phytate P contents. There was considerable variation found in most of the ingredient components. The average values for the laboratory determinations versus NIRS predictions were all within 0.030 for total phosphorus and 0.012 for phytate P. For phytate P, the magnitude of the standard errors of the predictions ranged from 0.009% for soybean meal to 0.012% for canola meal. These values may be sufficiently precise for nutritionists to use the NIRS predictions to estimate how much of the P in their ingredients is not available to the birds. For total P, the magnitude of the standard errors of the predictions ranged from 0.027% for corn DDGS to 0.142% for wheat middlings. In general, total P predictions by NIRS were not generally sufficiently precise for most nutritionists to use in feed formulation. Decision making may be quite easy in using NIRS estimates for the phytate P content of bakery by-product meal [R(2) = 0.89 for predicted = f (determined)] but not for the total P content of soybean meal (R(2) = 0.03). It is concluded that precise estimates of phytate P through NIRS should allow nutritionists for more efficient formulate and mix feed, lowering feed costs and reducing the amount of residual polluting phosphorus in poultry excreta.