Use this URL to cite or link to this record in EThOS:
Title: Utilisation of daily milk records in genetic evaluation of dairy cattle
Author: Olori, Victor Enishede
Awarding Body: University of Edinburgh
Current Institution: University of Edinburgh
Date of Award: 1997
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
The aim of this study was to determine how daily yield records might be used to improve the genetic evaluation of dairy cattle, which is currently based on 305-day yield predicted from monthly test day records. Daily milk yield records of 488 first lactation Holstein Friesian cows were obtained from one UK herd and summarised into weekly averages. Weekly fat, protein and lactose content records, from the same herd, were also obtained and each multiplied by test day yield to estimate fat, protein and lactose yields. Analysis of variance indicated that residual standard deviation (RSD) for each trait was lower when season of production was included in the model instead of the season of calving. The difference in RSD was more for fat and protein yields than the other traits. Lactation stage and season were the most important environmental factors affecting daily milk yield and composition while calving age had a small but significant (P<0.05) effect. Pregnancy accounted for 1.4 to 1.7% of the variation in yield traits but less than 0.4% of the variation in content traits. Its effects varied with gestation stage causing daily milk yield to decline by 3kg in the 8th month of gestation. A significant interaction between lactation and gestation stage was observed which suggested that the negative effect of pregnancy was higher in mid than late stages of lactation. Standard models of the lactation curve studied accounted for a substantial proportion of the variation in daily milk yields of typical lactations, which made up about half of the lactations studied. The other half was made up of atypical lactations such as highly persistent animals with almost flat curves whose lactation could not be adequately modelled with the standard curve functions. A regression spline model was derived which was as good as the best 3 parameter model and more flexible. These results indicate that models which make rigid assumptions about the shape of the lactation curve may not be very effective in accounting for the effect of lactation stage on daily yields. Yields of the same trait at different stages of lactation were positively correlated throughout but the correlation between yield and content traits was negative. Average correlation between milk yield and adjacent weeks was 0.93 declining to 0.61 between yields 41 weeks apart. Daily milk yields in mid and late stages of lactation were more highly correlated with 305-day yield than yields in early lactation. Covariance functions, using orthogonal polynomials up to the order of 4, were used to model genetic and permanent environmental covariances in a restricted maximum likelihood (REML) random regression (RR) model. Genetic parameters and breeding values were estimated for yield in every week of lactation. There were compared with estimates from a multivariate model, which considered yield at different stages as different traits, and a repeatability model without random regressions.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available