618
Mid-infrared Spectroscopy to Predict Important Milk Quality Traits in Irish Cows
The aim of this study was to test the effectiveness of mid-infrared spectroscopy (MIRS) to predict milk quality traits including free amino acids (FAA), milk protein fractions, and milk color. Milk samples were collected from a range of dairy breeds and individual milk proteins, FAA and color were quantified. Equations were developed to predict milk quality using a range of wavelengths of the MIR spectrum generated through routine milk recording as predictor variables. Prediction equations were tested using external validation. The correlation between true and predicted values (Rv) of protein fractions ranged from 0.27 (β-lactoglobulin A) to 0.56 (α-s2-casein), the Rv of milk color ranged from 0.00 (redness) to 0.71 (yellowness) while the Rv of FAA ranged from 0.12 (Threonine) to 0.72 (Valine).
Keywords:
Mid-infrared spectroscopy
Free amino acids
Protein composition