Feed analysis with near-infrared spectroscopy (NIRS)

Importance of the Topic

This application note addresses the critical need for rapid and reliable quality control of animal feed in livestock production. Traditional wet chemistry methods for determining moisture, protein, fiber, ash, starch, and fat content are time-consuming, labor-intensive, and require multiple reagents. Near-infrared spectroscopy (NIRS) offers a non-destructive, multi-component analysis in seconds, facilitating better feed management and ensuring safe, nutrient-balanced animal diets.

Objectives and Study Overview

The main objective was to evaluate the feasibility of using NIRS to predict key feed quality parameters across diverse feed types (poultry, swine, cattle, goats, sheep). Approximately 500 samples were collected and analyzed to build calibration and validation models for moisture, crude protein, crude fiber, ash, starch, and fat content.

Methodology

Ultrafine powdered or intact feed samples were placed in a rotating large-cup module and measured in reflectance mode to minimize inhomogeneity. Spectra were collected from multiple spots per sample and averaged. Calibration models were developed using multivariate regression, with cross-validation to assess predictive performance. Figures of merit (SEC, SECV, SEP, R²) were determined for each analyte.

Instrumentation

• Metrohm NIR Analyzer with rotating sample cup for reflectance measurements
• Integrated software for spectral acquisition, preprocessing, and chemometric model development

Main Results and Discussion

Calibration results demonstrated strong correlations between NIR predictions and reference methods:

• Fat: R² ≈ 0.92, SEC ~0.30%, SEP ~0.34%
• Starch: R² ≈ 0.93, SEC ~1.80%, SEP ~1.27%
• Protein: R² ≈ 0.87, SEC ~0.81%, SEP ~0.80%
• Moisture: R² ≈ 0.94, SEC ~0.41%, SEP ~0.31%
• Fiber: R² ≈ 0.65, SEC ~0.77%, SEP ~0.70%
• Ash: R² ≈ 0.72, SEC ~0.39%, SEP ~0.39%

Performance varied slightly by feed type but remained within acceptable precision for routine QC. Poultry and pig feed models achieved R² values above 0.90 for most parameters; cattle feed models were especially robust for starch and protein.

Benefits and Practical Applications

Implementing NIRS for feed analysis offers:

• Rapid, multi-component determination in under one minute per sample
• Elimination of chemical reagents and hazardous waste
• Reduced labor and operational costs
• On-site or at-line testing for real-time process control

Future Trends and Opportunities

Advancements likely to enhance NIRS feed analysis include:

• Portable and handheld NIR devices for field applications
• Integration with process automation and inline monitoring
• Improved chemometric algorithms and cloud-based calibration transfer
• Extension to novel feed ingredients (algae, insect protein) and contaminants screening

Conclusion

This study confirms that NIRS is a viable, efficient alternative to traditional wet chemistry for multi-parameter feed quality analysis. High correlation and precision support its adoption in feed mills and laboratories, offering significant time and cost savings while maintaining regulatory compliance.

References

• ISO 6493:2000 Animal feeding stuffs — Determination of starch content — Polarimetric method
• ISO 5984:2002 Animal feeding stuffs — Determination of crude ash
• ISO 6865:2000 Animal feeding stuffs — Determination of crude fibre content — Method with intermediate filtration
• ISO 5983:1997 Animal feeding stuffs — Determination of nitrogen content and calculation of crude protein content — Kjeldahl method
• ISO 6496:1999 Animal feeding stuffs — Determination of moisture and other volatile matter content
• ISO 6492:1999 Animal feeding stuffs — Determination of fat content