Analysis of Some Predictors for Estimating Carcass Weight in Brahman Cross Cattle

Abstract

Carcass weight is an important variable in beef cattle farm, noting that it can only be measured after being slaughtered. A method to estimate the carcass weight is needed it in order to reduce losses due to incorrect selection of cattle. The aim of this research is to analyze various variables as carcass weight predictors. The study was conducted at PT. KASA Central Lampung through a survey and field research from 145 cattle consisted of 80 heifers and 65 young steer. The data were collected by direct measurement in the field and added with the production records. The measured variables include carcass weight (CW), chest circumference (CC), body length (BL), body height (BH) and body condition score (BCS) which used as carcass weight predictors. All of the collected data were analyzed with linear and multiple regression to determine the predictors equation, and with one-way classification ANOVA with unbalanced model design by using Minitab software version 13.1. The results showed that sex significantly affect the carcass weight, and the predictors (CC, BL, BH and BCS) analysis showed that all predictors have very high accuracy (P<0.01) by using linear regression. Furthermore, the coefficient of determination (R2 adj. is 89.7% for CC, 71.2% for BL, 85.3% for BH, 71.6% for BCS) is lower than multiple regression in predicting the carcass weight (CW = - 470 + 0.748 CC + 0.878 BL + 2.23 BH + 15.5 BCS), with the coefficient of determination at R2 adj.>90%. The research concludes that the use of multiple regression with CW, CC, BL, BH, and BCS as predictors in the function would be more accurate compared to the linear regression with similar predictor variables.