본문 바로가기
HOME> 논문 > 논문 검색상세

학위논문 상세정보

임의회귀 모형을 이용한 젖소 산유형질의 유전모수 추정 원문보기
Estimation of genetic parameters for milk production traits in holstein using random regression test-day model

  • 저자

    김병우

  • 학위수여기관

    慶尙大學校 大學院

  • 학위구분

    국내박사

  • 학과

    응용생명과학부

  • 지도교수

  • 발행년도

    2003

  • 총페이지

    xvi, 144p.

  • 키워드

    임의회귀모형 젖소산유형질 유전모수 응용생명과학;

  • 언어

    kor

  • 원문 URL

    http://www.riss.kr/link?id=T10059998&outLink=K  

  • 초록

    The study was conducted to propose an alternative for the present national wide program for evaluating genetic ability of dairy cattle in Korea. The conventional method for evaluating lactation performance uses the production records fixed for 305 days. In the study, records of test-day were applied to analyze genetic parameters for milk production by test-day. Through the analysis, an efficacy of test-day Random Regression Model utilizing covariate functions was evaluated. The data of test-day in milk production and pedigree information of each individual were collected by National Agricultural Cooperative Federation according to the dairy herd improvement program and Korean Animal Improvement Association, respectively. The data for the analyses were from the animals in the first parity which have pedigree information and performance records more than three times in a lactation period. For the efficiency of data analysis, unusual records and animal records from the herd which was not connected in any blood relationships between herds were excluded. The applied Random Regression Model was a single trait animal model in which each lactation record was considered as an independent trait. Covariances of estimates were assumed to be different ones. The RRGibbs program(Meyer, 2002) was utilized and the Legendre polynomial covariate function was applied for the analysis. In the function, test-day was set up as a covariate and effect of age-season was considered as a fixed effect. In order to consider Heterogeneity of residual variance in the analysis using the RRGibbs program(Meyer, 2002), test-days were classified in 29 class in a lactation period. The REMLF90 program (Misztal, 2001) was applied when Heterogeneity of residual variance was not considered for the analysis. 1. From the result of the analysis considering Heterogeneity of residual variance, it is inferred that there is more individual variation for lactation performance in the early and late lactation periods which would be due to feeding management and physiological property of dairy cattle in Korea. After e6 that is the range between 61 and 70 test-day, there is less variation in residual variance and covariance. It indicates that there is more research should be performed to understand data property in the early lactation period when test-day model will be introduced for evaluating genetic performance of dairy cattle in Korea. However, consideration of Heterogeneity of residual variance could be excluded in a general way. 2. The heritability of milk yield by test-day was estimated between 0.154 ∼ 0.455 and the trend of a sudden increase of the heritability in the early lactation period (between the 20 and 25 test-day) was observed. Between the early and middle lactation period, there was only a little change of the heritability. After the 215 test-day, there was a slow decrease of the heritability. The highest estimate of heritability was h^(2)=0.455 at the 215 test-day and the lowest one was h^(2)=0.154 at the 305 test-day. The analysis showed that there was a broad variation in the heritability of milk yield by test-day. Most of the heritability estimated using the test-day model was higher than that (h^(2)=0.1461) using the Lactation Model. The low genetic variance (G) of milk yield estimated in the late lactation period indicates that individual variation in milk yield is more in the late lactation period. 3. The heritability of milk fat yield by test-day was estimated between 0.282 ∼ 0.517 and the trend of a sudden increase of the heritability in the early lactation period (between the 20 and 25 test-day) was observed. From the early to middle lactation period, there was only a little change of the heritability. After the 250 test-day, a slow increase of the heritability was shown. The highest estimate of heritability was h^(2)=0.517 at the 305 test-day and the lowest one was h^(2)=0.282 at the 20 test-day. The analysis showed that there was a broad variation in the heritability of milk fat yield by test-day. Most of the heritability estimated using the test-day model was higher than that (h^(2)=0.130) using the Lactation Model. 4. The heritability of milk protein yield by test-day was estimated between 0.263 ∼ 0.338 and the trend of an increase of the heritability in the early lactation period (between the 20 and 25 test-day) was observed. From the early to middle lactation period, there was only a little change of the heritability. After the 240 test-day, a slow decrease of the heritability was shown. The highest estimate of heritability was h^(2)=0.338 at the 20 test-day and the lowest one was h^(2)=0.263 at the 215 test-day. The analysis showed that there was a little variation in the heritability of milk protein yield by test-day. Most of the heritability estimated using the test-day model was higher than that (h^(2)=0.1153) using the Lactation Model. 5. The heritability of SNF yield by test-day was estimated between 0.088 ∼ 0.354 and the trend of a decrease of the heritability from the early lactation period, an increase from 45 days after parturition and another decrease from the 220 test-day was observed. The highest estimate of heritability was h^(2)=0.354 at the 220 test-day and the lowest one was h^(2)=0.088 at the 40 test-day. The analysis showed that there was a broad variation in the heritability of SNF yield by test-day. Most of the heritability estimated using the test-day model was higher than that (h^(2)=0.1153) using the Lactation Model. 6. The heritability of milk fat percentage by test-day was estimated between 0.145 ∼ 0.359 and the trend of a sudden increase of the heritability in the early lactation period (between the 20 and 25 test-day) was observed. From the 25 test-day to middle lactation period, the 230 test-day, there was a slow increase of the heritability. After the period, the heritability was decreased a little. The highest estimate of heritability was h^(2)=0.359 at the 230 test-day and the lowest one was h^(2)=0.145 at the 20 test-day. The analysis showed that there was a little variation in the heritability of milk fat percentage by test-day. Most of the heritability estimated using the test-day model was lower than that (h^(2)=0.4105) using the Lactation Model. 7. The heritability of milk protein percentage by test-day was estimated between 0.072 ∼ 0.497 and the trend of a sudden increase of the heritability at the begining of lactation period (between the 15 and 25 test-day) was observed. From the 15 test-day to middle lactation period, the 105 test-day, there was an increase of the heritability. From the 275 test-day, the heritability was decreased and then increased a little. The highest estimate of heritability was h^(2)=0.497 at the 105 test-day and the lowest one was h^(2)=0.072 at the 15 test-day. The analysis showed that there was a broad variation in the heritability of milk protein percentage by test-day. Most of the heritability estimated using the test-day model was lower than that (h^(2)=0.4264) using the Lactation Model. 8. The heritability of SNF percentage by test-day was estimated between 0.305 ∼ 0.489 and the trend of a sudden increase of the heritability at the begining of lactation period (between the 20 and 25 test-day) was observed. From the 25 test-day to middle lactation period, the 120 test-day, there was a slow increase of the heritability. From the 120 test-day, there was almost no change of the heritability. The highest estimate of heritability was h^(2)=0.489 at the 120 test-day and the lowest one was h^(2)=0.305 at the 20 test-day. The analysis showed that there was a variation in the heritability of SNF percentage by test-day only at the begining of lactation period (between the 20 and 25 test-day). Most of the heritability estimated using the test-day model was almost similar with that (h^(2)=0.4144) using the Lactation Model. 9. The estimates of heritability using 305 day corrected data were 0.146 for milk yield, 0.1300 for fat yield, 0.1153 for protein yield, 0.1125 for SNF yield, 0.4105 for fat percentage, 0.4264 for protein percentage and 0.4144 for SNF percentage. 10. The highest genetic correlation between milksum which is sum of EBV of each test-day and each EBV of test-day was at the 145 test-day (r=0.996). The highest genetic correlation between 305 day corrected EBV and each EBV of test-day was at 305 test-day (r=0.751). The genetic correlation between milksum and 305 day corrected EBV was estimated by 0.741. 11. The rank correlation coefficient for milk yield was estimated relatively high between test-days. The rank correlation coefficient between sum of EBV of each test-day and each 305 day corrected EBV was estimated by 0.712. This result indicates that the Random Regression Model could supplement the defect of Lactation Model using 305 day corrected data without considering the property of lactation stage. 12. The individuals holding the 1st to 8th positions in milk yield by the estimation using Lactation Model also exist within 50 individuals holding a high rank by the estimation using the Random Regression Model which considers Heterogeneity of residual variance. Moreover, 23 individuals were included within a high rank 50 individual estimated by both the Random Regression and Lactation Model. From the result, the Random Regression Model is thought to supplement the weak point of the Lactation Model in the evaluation of genetic performance for milk production of dairy cattle in Korea. 13. The comparison of Estimated Breeding Values by the Random Regression Model and Lactation Model also showed a similar result. Although a simple comparison of preponderance between the Random Regression Model and Lactation Model is not appropriate because they use different level of fixed effect and way of analysis, it is certain that the Random Regression Model can be an efficient alternative to evaluate genetic performance for milk production of dairy cattle in the future.


 활용도 분석

  • 상세보기

    amChart 영역
  • 원문보기

    amChart 영역