Improvement of accuracy in the estimation of lean meat content in pig carcasses

Abstract

Fat thickness including skin and muscle depth measured on the left carcass side between the second and third from the last rib 70 mm off the dorsal midline were measured in a total of 168 pig hybrid carcasses. The lean meat content was then determined on the basis of simplified dissections of the carcasses. Multiple regressions of the measurements of the fat and muscle thickness on the lean meat content obtained by dissections were used to construct the following basic regression formulae for the ultrasound and probe apparatuses: &gamma;_IS-D-05 = 60.69798 – 0.89211S_IS-D-05 + 0.10560M_IS-D-05 and &gamma; _IS-D-15 = 60.92452 – 0.77248S_IS-D-15 + 0.11329M_IS-D-15, respectively. To increase the accuracy of the prediction formulae, additional measures were included in the calculation which reduced s_e by 0.48 to 0.54 percent points. The relationships between the lean meat content and other indicators of carcass value were also assessed. The highest correlation coefficient was determined in the ratio of the fat cover area above the <I>musculus longissimus lumborum et thoracis</I> (MLLT) to the MLLT area (<I>r</I> = –0.87). On the contrary, the lean meat content demonstrated the lowest correlation with the cold carcass weight (<I></I>r = –0.25). Major carcass cuts (ham, loin, shoulder, belly with bones) from the carcasses classified in different SEUROP classes were evaluated. Significant differences between the classes were found in the proportions of cuts without fat cover, fat thickness measured at point “P₂”, and fat thickness measured on the midline plane separating the left and right sides of the carcass.