Département PHASE, INRA, VetAgro Sup, UMR Herbivores, Université Clermont Auvergne, Saint-Genès-Champanelle, France.
J Sci Food Agric. 2019 Jun;99(8):4129-4141. doi: 10.1002/jsfa.9643. Epub 2019 Mar 18.
This study implemented a holistic approach based on the farm-to-fork data at the four levels of the continuum (farm - slaughterhouse - muscle - meat) to study the inter-individual cluster variability of beef tenderness. For that, 171 young bulls were selected on a large database of 480 animals according to the industrial expectations based on animal and carcass characteristics. The targeted factors were age at slaughter (14; 20 months), carcass weight (370; 470 kg), EUROP conformation (7; 15) and fatness (2.5; 5) scores of the carcasses. Multivariate analyses and unsupervised learning tools were performed.
Principal component analysis combined to agglomerative hierarchical clustering allowed ten clusters to be identified that differed (P < 0.0001) for the four targeted factors. The clusters were further different for variables belonging to each level of the continuum. The results indicated an inter-individual cluster variability rising in tenderness in link with the continuum data grouped according to industrial expectations. The associations of the whole variables of the continuum with tenderness were very important, but farm-to-fork continuum-levels dependent. The findings showed that the variables contributing most to the inter-individual cluster variability of tenderness seemed to be more related to the rearing practices, mainly feeding, and their consequences on carcass properties rather than to the muscle characteristics evaluated by enzyme metabolism and connective tissue.
It seems that considering the continuum data would allow possible trade-off managements of tenderness to identify levers at different levels from the farm-to-meat. © 2019 Society of Chemical Industry.
本研究基于从农场到餐桌的数据在连续体的四个水平(农场-屠宰场-肌肉-肉)实施整体方法,研究牛肉嫩度的个体间聚类变异性。为此,根据动物和胴体特性的工业预期,从 480 只动物的大型数据库中选择了 171 头年轻公牛。目标因素为屠宰时的年龄(14;20 个月)、胴体重量(370;470 公斤)、欧盟体型(7;15)和脂肪(2.5;5)评分。进行了多元分析和无监督学习工具。
主成分分析与凝聚层次聚类相结合,确定了 10 个不同的聚类(P<0.0001),四个目标因素存在差异。根据工业预期分组的连续体数据,聚类在各个级别上的变量也存在差异。结果表明,嫩度的个体间聚类变异性随着与工业预期相符的连续体数据的增加而增加。连续体与嫩度的整体变量的关联非常重要,但取决于连续体水平。研究结果表明,对嫩度的个体间聚类变异性贡献最大的变量似乎与饲养实践,主要是饲养,及其对胴体特性的影响更相关,而不是通过酶代谢和结缔组织评估的肌肉特性。
似乎考虑连续体数据可以允许在从农场到肉的不同层面上进行可能的权衡管理,以确定嫩度的杠杆。© 2019 化学工业协会。