On-line application of visible and near infrared reflectance spectroscopy to predict chemical–physical and sensory characteristics of beef quality

The aim of this study was to assess the on-line implementation of visible and near infrared reflectance (Vis-NIR) spectroscopy as an early predictor of beef quality traits, by direct application of a fibre-optic probe to the muscle immediately after exposing the meat surface in the abattoir. Samples from M.longissimus thoracis from 194 heifers and steers were scanned at quartering 48 h postmortem over the Vis-NIR spectral range from 350 to 1800 nm. Thereafter, samples from M.longissimus thoraciset lumborum were analysed for colour (L∗, a∗, b∗; 48 h postmortem), cooking loss (14 days postmortem), instrumental texture (Volodkevitch, 10 days aged meat; slice shear force, 3 and 14 days aged meat) and sensory characteristics. Vis-NIR calibrations, tested by cross-validation, showed high predictability for L∗, a∗ and b∗ (R2 = 0.86, 0.86 and 0.91; SECV = 0.96, 0.95 and 0.69, respectively). The accuracy of Vis-NIR to estimate cooking loss and instrumental texture ranged from R2 = 0.31 to 0.54, suggesting relatively low prediction ability. Sensory characteristics assessed on 14 days aged meat samples showed R2 in the range from 0.21 (juiciness) to 0.59 (flavour). Considering the subjective assessment of sensory characteristics the correlations of Vis-NIR measurements and several meat quality traits in the range from 0.46 to 0.95 support the use of on-line Vis-NIR in the abattoir. Improvement of predictability was achieved if only extreme classes of meat characteristics have to be predicted by Vis-NIR spectroscopy.