Selection for body weight in dairy cattle

This thesis deals with selection for body weight (BW) in dairy cattle. The economic efficiency of present breeding schemes might increase further when selection decisions also consider information on BW as BW relates to feed costs and revenues from beef production. However, the practical implementation of such a selection strategy is hindered by limited knowledge on procedures for data recording and genetic evaluation. The aim of this thesis was to study the possibilities and economic relevance of selection strategies for BW in dairy cattle. To obtain information on the optimal moment and frequency of data recording, this thesis analysed repeated BW observations on growing and lactating heifers from experimental herds. A study on growth patterns of heifers from birth until first calving fitted by two mathematical functions suggested that BW at first calving has a high correlation with mature BW.Repeated observations for BW were used to estimate non-genetic effects and repeatabilities during the first three parities. The most important non-genetic effects on BW were feeding, lactation stage, pregnancy stage and age at calving. Observations for BW were highly repeatable within and across parities. Observations at any lactation stage were highly correlated with mean BW during lactation. These results indicated that a single BW observation on a lactating heifer is a good predictor of mean BW during the first three parities.A genetic standard deviation of 29 kg and a heritability of 0.33 were estimated from a large field experiment. The high correlations between BW and conformation scores show that a large-scale genetic evaluation for BW can easily be based on conformation data that are already routinely recorded in most breeding schemes. A study of the dynamic aspects of BW during lactation showed that it is possible to select for mean BW during lactation but that opportunities to select for specific BW change patterns were limited.This thesis also estimated the genetic parameters of BW and DMI during lactation from repeated observations using covariance functions methodology. Heritability estimates for BCS and DMIC during lactation were in the range 0.21-0.45 and 0.18-0.37, respectively. The genetic background was found to be constant during lactation for BCS but not for DMIC. Heritability estimates for BCS from routinely collected BCS data were around 0.35.A farm model was used to estimate economic values (EVs) for BW and DMIC under present and future production circumstances. The present EV for BW was negative and might become more negative under future production circumstances with lower beef prices and higher milk production levels. The present EV for DMIC was positive and might become more positive under more extensive production circumstances with higher milk production levels.Finally, several alternative selection strategies using EBVs for BW were evaluated. The inclusion of EBVs for BW without accounting for changes in DMIC reduced the positive trends in BW, milk production and DMIC but increased the unfavourable trends in BCS and calving interval. The additional economic response of this alternative was small (+0.4%). When also accounting for changes in DMIC, the optimal weighting for EBVs for BW in the index became positive. In this situation, genetic trends for milk production, BW and DMIC increased. The economic response increased by 0.9%. However, the main disadvantage of this alternative is the unfavourable effect on durability and calving interval. Despite the negative weighting on BW in the index, the positive genetic trend for BW was even higher than in the present selection strategy. The responses of this index were highly sensitive to the EVs and the reliability of EBVs for DMIC.It was concluded that despite a negative weighting of BW in the breeding goal, a negative weighting of EBVs for BW cannot be recommended under present Dutch production circumstances.