FT-NIR spectroscopy of fresh and treated muscle tissue in young female rabbits

A total of 32 young female rabbits (live weight 3.648 g) were made to fast or allowed to eat for 36 h before slaughter. Samples of the Longissimus lumborum (LL) muscle and two strips of the Obliquus abdominis muscle (OA) were sampled from chilled carcasses. The muscles were divided into three parts: 1) to determine the moisture; 2) placed in a plastic tube, then immersed in 95% ethanol; 3) stored at -18 °C, then freeze-dried. A further sample, from the dissected hindleg (HL), was stored at -18 °C and freeze-dried. The quality traits were measured in raw LL, cooked or alcohol treated. An interesting close relationship (r=0.78) emerged between the two substrates used for shear force measurements: alcohol vs. cooked. The muscles were submitted to FT-NIR spectroscopy as fresh tissue (LL, OA and HL), then as intact freeze -dried (LL_if and HL_if) and also as ground freeze-dried (LL_gf, OA_gf and HL_gf), and were then used to predict the intramuscular lipid content. Two intact alcohol samples (LL_e and OA_e) were examined. The spectra were correlated and cross-validated to fixed experimental effects as binary data and to quantitative traits. 1-VR cross-validated values were reported. The fasting effect shown by the NIR was almost four times lower than the replication effect (av.ge 1-VR = 0.12 vs 0.52). The freeze-drying of substrates amplified the effects in the NIR spectra, while the alcohol treatment generally decreased them. The estimation of the intramuscular lipid content through scanning of the intact muscles was nearly half-efficient vs scanning of the ground freeze-dried tissue for OA (0.55) and for HL (0.53), but was poor for LL (0.16). An agreement between the spectra of OA and the presence of noising and experimental effects was confirmed. The LL in ethanol, which is easy to transport to a laboratory, seems to be suitable for tenderness measurements, but the NIR scan should be further investigated for meat quality assessment.