Análise multivariada de características de qualidade de carcaça e da carne de bovinos Angus-Nelore terminados em confinamento
Carregando...
Data
2019-05-15
Autores
Lopes, Lucas Silva Ferlin
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Estadual Paulista (Unesp)
Resumo
A produção brasileira apresenta alta variabilidade na qualidade da carcaça e da carne bovina. Ao ser influenciada por todos os elos da cadeia produtiva, a qualidade final da carne precisa ser avaliada e assegurada. Análises estatísticas multivariadas como a análise de componentes principais (PCA) e métodos de agrupamento não-hierárquico como o k-means configuram opções para compreender as interações entre as características relacionadas à qualidade. O presente estudo teve como objetivo avaliar características de qualidade de carcaça e de carne de novilhos ½ Angus x ½ Nelore terminados em confinamento por meio da PCA. Bem como realizar a separação dos animais por análise de agrupamento não hierárquico k-means. Os quatro primeiros componentes principais (PC) explicaram aproximadamente 58% da variabilidade total e as variáveis de cor da carne foram mais efetivas para definir o primeiro PC. Peso de carcaça e retenção de água definiram o segundo, rendimento de carcaça definiu o terceiro e o teor de gordura definiu o quarto PC. Na análise de agrupamento, três grupos de chroma foram formados e projetados no plano definido por PC1 e PC2. As médias de chroma nos grupos foram 18,07 ± 1,90; 22,75 ± 1,15; 26,08 ± 0,79; com 25, 48 e 24 animais em cada grupo, respectivamente. O aumento para chroma nos grupos foi acompanhado por maiores valores de PCQ, PCF, COL, PRT, PPE, PT, e por menores valores de PPG, FC e UMD. Portanto, a PCA e k-means mostraram que a cor pode ser utilizada como ferramenta de identificação de animais e de carnes com qualidade superior.
Brazilian production presents high variability in carcass and beef quality. Influenced by all links in the production chain, the final quality of the meat needs to be evaluated and ensured. Multivariate statistical analyzes such as principal component analysis (PCA) and non-hierarchical clustering methods such as k-means configure options to understand the interactions between quality related traits. The aim of the present study was to evaluate carcass and meat quality traits from ½ Angus x ½ Nelore steers finished in feedlot, trough PCA. As well as performing the clustering of the animals trough non-hierarchical k-means cluster analysis. The first four principal components (PC) explained roughly 58% of the total variability and meat color variables were most effective in defining the first PC. Carcass weight and water holding capacity defined the second, carcass yield defined the third and the fat content defined the fourth PC. In cluster analysis, three chroma groups were formed and projected in the space defined by PC1 and PC2. Chroma mean values in the groups were 18.07 ± 1.90; 22.75 ± 1.15; 26.08 ± 0.79; with 25, 48 and 24 animals in each group, respectively. The increase for chroma in the groups was followed by higher values of PCQ, PCF, COL, PRT, PPE, PT, and by lower values of PPG, FC and UMD. Therefore, PCA and k-means have shown that color can be used as a tool to identify animals and meat of higher quality
Brazilian production presents high variability in carcass and beef quality. Influenced by all links in the production chain, the final quality of the meat needs to be evaluated and ensured. Multivariate statistical analyzes such as principal component analysis (PCA) and non-hierarchical clustering methods such as k-means configure options to understand the interactions between quality related traits. The aim of the present study was to evaluate carcass and meat quality traits from ½ Angus x ½ Nelore steers finished in feedlot, trough PCA. As well as performing the clustering of the animals trough non-hierarchical k-means cluster analysis. The first four principal components (PC) explained roughly 58% of the total variability and meat color variables were most effective in defining the first PC. Carcass weight and water holding capacity defined the second, carcass yield defined the third and the fat content defined the fourth PC. In cluster analysis, three chroma groups were formed and projected in the space defined by PC1 and PC2. Chroma mean values in the groups were 18.07 ± 1.90; 22.75 ± 1.15; 26.08 ± 0.79; with 25, 48 and 24 animals in each group, respectively. The increase for chroma in the groups was followed by higher values of PCQ, PCF, COL, PRT, PPE, PT, and by lower values of PPG, FC and UMD. Therefore, PCA and k-means have shown that color can be used as a tool to identify animals and meat of higher quality
Descrição
Palavras-chave
Análise multivariada, Bovino de corte, Carne qualidade