Epsilon Open Archive

Abstract

Reproductive efficiency in dairy cows is a key factor for milk producers, and numerous studies have identified impaired reproductive performance as a major cause of reduced production efficiency in the dairy industry. The overall aim of this thesis was to gain knowledge of factors affecting the reproductive performance indicators currently used by herd advisory services and to find other, possibly more efficient, ways to measure reproductive performance in dairy cows.

The studies included in this thesis were based mainly on records from the Cattle Database at the Swedish Dairy Association, but also on data from a simulation model. The records used in the constituent studies of the thesis came from 2728 herds (Paper I), 483 herds (Paper II), 900 simulated herds (Paper III), and 132,721 individual cows (Paper IV). The statistical analyses were performed using multivariable linear and logistic regression models as well as survival analysis.

Many statistically significant associations were found between herd and cow characteristics and reproductive performance. When allocating advisory service resources to improve reproductive performance, the focus should be on easily influenced herd characteristics, such as heat detection efficiency, nutritional status, do-it-yourself inseminations, and health deviations. Herd characteristics were also found to influence the degree of disagreements in reproductive performance indicators when data were updated after six months. Reproductive performance indicators that were adjusted for management strategy were evaluated, and percentage pregnant after the herd voluntary waiting period plus 30 days was found to be the best reproductive performance indicator with which to assess both reproductive management and reproductive physiology. Both percentage pregnant after the herd voluntary waiting period plus 30 days and percentage inseminated after the herd voluntary waiting period plus 30 days could be integrated for use in the herd advisory services. These indicators would be useful in providing information on herd performance and in establishing more efficient benchmarking between herds.