Estimation of meat amount by non-linear multiple regression equations using in vivo and carcass measurements on Teleorman Black Head lambs

In the present study non-linear multiple regression equations and carcass ultrasound measurements were used to estimate the amount of meat in carcass and commercial cuts in local breed Teleorman Black Head (TBH). The measurements were conducted on 79 TBH lambs aged 2.5 months, in two points (P1 - located 5 cm from the spine, in line with the 12th rib; P2 - located between 3rd and 4th lumbar) of longisimus dorsi muscle to obtain the following parameters: subcutaneous fat layer thickness (2.21; 2.03 mm), muscle depth (20.81; 19.54 mm), muscle eye area (8.93; 8.71 cm(2)) and muscle perimeter (121.97; 121.57 mm). The non-linear multiple regression equations based on all four ultrasound parameters measured in P1 gave the most precise estimations for carcass meat and commercial cuts: leg and loin (0.994), shoulder (0.963) and rack (0.938). The best estimations of the carcass meat amount and half carcass meat amount using three ultrasound parameters (depth, eye area and perimeter of muscle) were obtained in P1, with a precision of 0.818 and 0.803, respectively. Non-linear multiple regression equations using only one ultrasound parameter (muscle eye area) measured in P2 gave the most precise estimations for: carcass meat (0.916), half carcass meat (0.880) and commercial cuts such as loin (0.976), rack (0.950) and shoulder (0.911). The non-linear multiple regression equations developed by using ultrasounds parameters showed very high precision coefficients, which suggests that only ultrasound measurements and proposed equations might be used to estimate the meat production and to improve the evaluation of sheep selected for meat production.