Feed analysis with near-infrared spectroscopy (NIRS)

Livestock production is a complex process involving many independent and integrated operations. According to the World Health Organization (WHO), the foundations for successful livestock production are the availability of animal feed and efficient feeding. Quality control (QC) of feed and feed ingredients is crucial to ensure the production of safe and nutrient-rich animal feed. Typical QC parameters measured during the production of feed are fat, moisture, protein, fiber, ash, and starch. However, analyzing these parameters is usually done with wet chemistry, requiring different time-consuming procedures, complex analysis methods, and costly chemical reagents. Near-infrared spectroscopy (NIRS) offers rapid and reliable prediction of fat, moisture, protein, fiber, ash, and starch content of different feed types in a few seconds without any sample preparation.

Around 500 samples of feed (e.g., for poultry, swine, cattle, goats, and sheep) were analyzed on a Metrohm NIR Analyzer. All measurements were performed in reflection mode using the large cup. The samples were measured in rotation to collect spectral data from several areas. Spectral averaging of signals from various spots helped to reduce influences from sample inhomogeneity. Metrohm software was used for all data acquisition and prediction model development.

All measured NIR spectra were used to create prediction models for quantification of the key quality parameters for feed.

In addition, prediction models for moisture, starch, fiber, protein, fat, and ash were created for individual feeds, e.g., cattle feed, poultry feed, and pig feed.

The quality of the prediction models was evaluated using correlation diagrams (Figures 2–7) which display a very high correlation between the NIR prediction and the reference values. The respective figures of merit (FOM) display the expected precision of a prediction during routine analysis of NIR feed testing (Table 1 poultry feed, Table 2 swine feed, and Table 3 cattle feed).

Figures of merit

The following tables display the figures of merit for the prediction models of specific feed products: poultry feed (Table 1), pig feed (Table 2), and cattle feed (Table 3).