Prediction of beef quality traits through mini NIR spectrophotometer and multivariate analyses

Abstract

The aim of this study was to test the ability of mini NIR reflectance spectroscopy to predict beef quality traits. Sixty M. longissimus thoracis were collected and spectra were obtained prior to beef quality trait analysis. Calibration equations were developed from reference data (n=60) of pH, color traits (lightness, redness and yellowness), drip loss (%), cooking loss (%), CP (%), EE (%), moisture (%), DM (%), and Ash (%) using partial least squares regressions. Predictive ability of the models was assessed by coefficient of determination of cross-validation (R2CV) and root mean square error of cross-validation. Predictions models were satisfactory (R2CV = 0.95) for pH, (R2CV = 0.96) for lightness (L*), (R2CV = 0.96) for redness (a*), (R2CV = 0.97) for yellowness (b*), (R2CV = 0.95) for drip loss, (R2CV = 0.95) for cooking loss, (R2CV = 0.94) for CP, (R2CV = 0.95) for EE, (R2CV = 0.91) for moisture, (R2CV = 0.91) for DM and (R2CV = 0.91) for ash. The ratio performance deviation is 5.35, 5.34, 5.87, 5.16, 4.64, 4.81, 4.45, 4.95, 3.36, 4.73 and 4.47 for L*, a*, b*, pH, drip loss, cooking loss, CP, EE, moisture, DM and Ash respectively which indicates that all values are adequate for analytical purposes. Range error ratio are 20.69, 22.97, 27.11, 18.92, 20.74, 16.20, 17.80, 17.52, 14.96, 17.89 and 17.87 for L*, a*, b*, pH, drip loss, cooking loss, CP, EE, moisture, DM and ash respectively. From the findings of this study it can be concluded that mini NIRS is a suitable tool for a rapid, non-destructive and reliable prediction of beef quality.