Research of Statistical Model for Genetic Evaluation of Hanwoo Carcass Traits

2006년부터 2009년까지 축산물품질평가원에서 등급 판정된 도체성적이 있는 개체와 한국종축개량협회에 등록이 확인된 231,382 두의 자료를 이용하여 각 도체형질의 자료구조 분석 등을 살펴보고, 각 요인별로 도체형질에 어떠한 영향을 주는 지에 대한 환경효과 등을 분석하고, 변수선택법 등을 이용하여 향후 유전능력평가에서 사용 가능한 통계모형 찾아내고자 실시하였다. 한우 암, 수, 거세를 포함한 전체집단에서 출하체중, 도체율, 등 지방두께, 근내지방도에서 정규분포를 보이고 있었고, 다른 형질은 정규분포에 근접함을 확인할 수 있었다. 각 요인에 대한 분산분석에서는 한우 도체형질에 대하여 성별, 출생년도-계절, 도축년도-계절, 출생지역, 도축일령의 모두 고도의 유의성(p<0.01)이 인정되었다. 변수선택법을 이용한 모형적합도 검정에서는 출하체중, 도체중, 도체율, 배최장근단면적은 다섯 가지 환경효과를 선택하는 것이 최적이었고, 등지방두께와 근내지방도는 네 가지 환경효과를 고려하는 것이 최적이었지만, 다형질 분석으로 한 번에 분석을 하기 위해서는 다섯 가지 환경효과를 함께 고려하는 것이 적합하다고 사료된다. This study was conducted to study the environment effects on live weight, carcass weight, dressing percentage, eye muscle area, backfat thickness, and marbling score, which are the carcass traits of Hanwoo, based on the estimates and all the possible regression for the selection of variable and significance test for 231,382 heads that underwent the carcass measurements. The average and standard deviation for the live weight, carcass weight, dressing percentage, eye muscle area, backfat thickness, and marbling score were 654.79${\pm}$91.61 kg, 362.30${\pm}$67.15 kg, 59.52${\pm}$0.03%, 81.79${\pm}$12.21 $cm^2$, 11.39${\pm}$5.40 mm, 4.38${\pm}$2.29, respectively. The live weight, carcass weight, dressing percentage, eye muscle area, backfat thickness, and marbling score for cow were 532.79${\pm}$78.38 kg, 313.40${\pm}$44.90 kg, 56.50${\pm}$0.03%, 75.24${\pm}$10.69 $cm^2$, 11.82${\pm}$5.10 mm, 4.30${\pm}$2.06, respectively, while for bull were 619.74${\pm}$93.27 kg, 376.89${\pm}$48.62 kg, 58.61${\pm}$0.02%, 85.61${\pm}$10.46 $cm^2$, 5.64${\pm}$2.71 mm, 1.41${\pm}$0.83, respectively, and for steer were 681.78${\pm}$70.72 kg, 415.23${\pm}$49.43 kg, 60.19${\pm}$0.02%, 88.29${\pm}$10.27 $cm^2$, 12.71${\pm}$5.23 mm, 5.42${\pm}$1.99, respectively. In the environmental variables selection based on the variables selection method, the examination by carcass traits suggested that the most appropriate model could be determined when five variables were selected for the live weight, carcass weight, dressing percentage, eye muscle area, and four variables for backfat thickness, and marbling score. When they were considered at a time altogether based on multiple traits, it was deemed to be desirable to insert all five variables into the variables for analysis. In addition, high significance was found by carcass traits.