Multiple regression equations to estimate the content of breast muscles, meat, and fat in Muscovy ducks

Abstract

The aim of the present study was to derive multiple regression equations for in vivo estimation of the carcass lean and fat content in Muscovy ducks. The experimental materials consisted of 240 White Muscovy ducklings (120 ♂ and 120 ♀). One hundred sixteen females aged 10 wk and 112 males aged 12 wk were slaughtered. Before slaughter the ducks were weighed, and the following body measurements were taken: humerus length, drumstick length, chest girth, breast-bone crest length, width between the humeral bones, chest depth, and breast muscle thickness. The coefficients of simple correlation between carcass tissue components and body measurements were calculated. It was found that live body weight was highly correlated with the weights of all tissue components (r = 0.701 to 0.857). In males a significant interrelation was found between breast muscle weight and all body measurements, whereas in females breast muscle weight was correlated with breast-bone crest length, chest girth, width between the humeral bones, chest depth, and breast muscle thickness only. In both males and females the carcass lean content was closely correlated with drumstick length, breast-bone crest length, chest girth, and width between the humeral bones. In drakes the carcass fat content was closely correlated with all body measurements, whereas in hens significant correlations were observed between the carcass fat content and chest girth, width between the humeral bones, and chest depth only. The coefficients of simple correlation between the percentages of carcass tissue components and body measurements were generally low and statistically nonsignificant. Twelve multiple regression equations formulated based on the body measurements of live ducks were verified with respect to the accuracy of estimation of the content of breast muscles, meat, and fat with skin in the carcass. These equations give small SE of the estimate (Sy = 23.3 to 83.8 g), high values of coefficients of multiple correlation between the dependent variable and the set of independent variables, and high values of determination coefficients.