Use of NIRS for the assessment of meat quality traits in open-air free-range Iberian pigs

Abstract Near infrared spectroscopy (NIRS) can be useful in order to determine meat quality traits as a rapid and non-destructive technique. The aim of the present study is to assess the accuracy of NIRS technique to determine meat quality traits on Longissimus thoracis et lumborum muscle of open-air free-range Iberian pig (n = 287) both in intact and minced samples. Traits were measured by instrumental-chemical techniques: colour (L*, a* and b*), myoglobin content, centrifuge force water loss and texture: shear force and texture profile analysis (TPA: hardness, cohesiveness, springiness, chewiness). Calibration models between instrumental-chemical measures and NIRS spectral data were developed employing partial least square regressions (PLS). The samples were split in two random datasets (80 % in training set, 20 % in external validation set). An internal full cross-validation method was applied. Results were evaluated in terms of coefficient of determination (R2), root-mean-square error (RMSE), residual prediction deviation (RPD) and range error ratio (RER). Full spectral range was used to develop mathematical equations. As regard external validation procedure, the highest coefficients of determination (R2p) in intact loin samples were achieved for hardness, redness (a*) and yellowness (b*) (0.7 0.7. The models developed suggest the ability of NIRS for quantitative prediction of shear force and for a rough classification of colour parameters L*, a*, b* in minced samples as well as hardness in intact samples.