Infrared milk analyzers: Milk urea nitrogen calibration.

Area of Science:

• Dairy Science
• Analytical Chemistry
• Spectroscopy

Background:

• Accurate measurement of milk urea nitrogen (MUN) is crucial for dairy herd management and nutritional strategies.
• Existing calibration sample sets may not adequately address variations in other milk components, affecting MUN prediction accuracy.
• Mid-infrared (MIR) spectroscopy coupled with partial least squares (PLS) regression is a common method for MUN prediction, but its performance can be influenced by sample matrix effects.

Purpose of the Study:

• To redesign a milk calibration sample set for a wider, well-distributed range of milk urea nitrogen (MUN) concentrations.
• To ensure the new calibration set is orthogonal to variations in fat, protein, and lactose.
• To evaluate and enhance the performance of MIR-PLS models for MUN prediction using the redesigned calibration set.

Main Methods:

• Modified a 14-sample milk calibration set formulation and manufacturing process.
• Determined MUN reference values using a spectrophotometric enzymatic method.
• Calibrated three MIR milk analyzers with the modified samples and evaluated PLS model performance.

Main Results:

• The redesigned calibration set achieved a desired range of MUN concentrations.
• Within- and between-laboratory variations for MUN reference values were 0.43% and 0.77%, respectively.
• Model performance improved by 50% (reduced standard deviation of differences between predicted and reference MUN values) after applying intercorrection factors.

Conclusions:

• The redesigned, orthogonal milk calibration set effectively supports accurate MUN reference value determination.
• The developed method significantly improves the performance and reliability of MIR-PLS models for MUN prediction in milk.
• This approach enhances the accuracy of MUN measurements, benefiting dairy nutrition and management.