Predictive model on tenderness of postmortem yak meat and its validation

Tenderness is one of important sensorial properties for the quantitative evaluation of eating qualities influencing the consumption of beef, and the others are juiciness and flavor respectively. Poor tenderness is one of main causes for commercial unacceptability of yak meat. The variation of physiological and biochemical characteristics during tenderization at the initial phase of postmortem aging has an extraordinarily significant effect on the postmortem formation of the tenderness property of beef cuts, which is one of determinant factors for consumers to decide if the beef products should be purchased. Therefore, the improvement of research on the effect of physiological and biochemical characteristics on the tenderness of yak meat is beneficial to melioration of the theory of postmortem aging and tenderization of yak meat. In order to investigate the effect of the postmortem variation of physiological and biochemical characteristics on the formation of tenderness of yak meat, pH, glycogen content, calpain activity (CA), the myofibrillar fragmentation index (MFI), sarcomere length, myofiber diameter, and shearing force of M. longissimus dorsi stored at 4°C for 10d from 18 male yaks were determined. The estimated value of 6 characteristics at postmortem 1 h were obtained directly as alternative predictors of a predictive model of postmortem tenderness of yak meat. Meanwhile, the kinetic parameters of the rate and extent of variation of some physiological and biochemical characteristics including pH, CA, MFI and glycogen content obtained by kinetic data analysis within postmortem 72 h were also selected as alternative predictors. The ultimate predictors of a predictive model of postmortem tenderness of yak meat were determined through a jackknife method from all alternative predictors, including 6 variables at postmortem 1 h and 7 kinetic parameters within postmortem 72 h according to occurrences in optimal equations of 15 jackknife samples through multiple linear step regression analysis. The ultimate predictors included rate of CA decline (RCA), extent of pH decline (EpH), rate of pH decline (RpH), and rate of glucogen decline (RGlu). The predictive model of tenderness was also built within a randomly selected 15 male yaks different from the jackknife samples through multiple linear enter regression analysis, to shearing force at postmortem 6 d (T6) as the dependent variables, to the above four predictors as independent variables. The predictive equation is T6 = 50.414 + 10.924EpH + 53.029RpH-519.690RCA + 65.757RGlu, with determinate coefficient R2 of 0.703, which could explain 70.3% of the variability of the shearing force of yak meat at postmortem 6d. The data of the other 3 male yaks was used as a validation set. The determinate coefficient R2 between the predictive value obtained by model and reference value of the validation set was 0.977. The relative errors between predictive value and reference value of the validation set were not higher than 5%. Therefore, the tenderness of yak meat at postmortem 6d could be effectively predicted through this predictive model, based on the variation of physiological and biochemical characteristics within postmortem 72h.