Prediction of carcass tissues composition using the neck and shoulder traits in hair lambs with multiresponse multivariate adaptive regression splines

Abstract

The present study aimed to predict the carcass tissue composition of hair sheep lambs using a multiresponse multivariate adaptive regression splines algorithm. The left half of sixty-six hair lambs were divided into five commercial cuts (neck, shoulder, rib, loin, and leg), each cut was weighed and dissected in total soft tissue (fat and muscle, TSTW) and bone (BOW). The independent variables included variables obtained from neck and shoulder dissection: weights of the neck (NWE) and shoulder (SWE), neck soft STW (NSTW), neck BW (NBOW), shoulder STW (SSTW), and shoulder BW (SBOW). The prediction of hot carcass weight (HCW), cold carcass weight (CCW), carcass soft tissue weight (CSTW), and carcass bone weight (CBWE) had an R2 that ranged from 0.90 to 0.96. It is concluded that some neck traits and all shoulder traits could be used to predict the carcass tissue weights of hair-suckling lambs correctly.