Analys av juverhälsodata från kärnbesättningen Viken

Abstract: Nötcenter Viken is a Swedish nucleus herd for dairy cattle. The goal for the farm is to select suitable mothers for the next top generation of bulls. Today the selection for udder health is based on a pedigree index and phenotype data. The pedigree index is based on the cow's father and grandfathers breeding values. In this study breeding values have been estimated with animal models which also account for the cows own performance. Breeding values in Sweden today are based on lactation averages. To be able to estimate a breeding value for a cow she needs to have a complete 305 days lactation. A mean value that represents the whole lactation is used to estimate a breeding value. An alternative to this method is to estimate breeding values with a test-day model. With a test-day model a breeding value can be estimated without registrations from a whole lactation. The aim of this study was to estimate breeding values with these two models for somatic cell count. A correlation with pedigree-index was also compiled. One other important part of this study was to see how available data collected at Nötcenter Viken was and how the data could be used in further research. The aim of the literature review was to survey different ways to measure and improve udder health in dairy cattle. Most common today is to use mastitis registrations, somatic cell count and udder type traits. Other possible measurements of udder health are electric conductivity in the milk and milking speed. All data were collected from Vikendata which is an on-farm-recording data system used at Nötcenter Viken. From Vikendata information about: pedigree, calving, cell counts, udder type trait, health and milking speed were collected. Half of the cows in this study were Swedish Red and the other half were Swedish Holstein. All cows were born after 2002 and had registrations for somatic cell count in their first lactation. Only 261 cows had a first lactation longer than 90 days. For these cows breeding values were estimated with the lactation average model and test-day model. An additional 45 cows had registrations for somatic cell count in their first lactation but for different reasons their lactation were shorter than 90 days. These 45 cows only obtained breeding values estimated with test-day model. The geometric average somatic cell count for cows in Sweden 2005/2006 was 183 000 cells/ml milk. In this study the geometric average for the 306 cows participating was 91 000 cells/ml. Effects of mastitis, lactation, number and calving season were all significant factors. Cows that have had mastitis at some point during a lactation had significantly higher cell counts than those that have not had mastitis. Calving during the summer resulted in moderately lower cell counts compared to other calving seasons. It is important to note that this study did not take into consideration breed, calving year or age at calving. The correlation between the two estimated breeding values were high: 0.92. Both estimated breeding values had a much lower correlation with the pedigree index -0.16 for breeding value estimated with the lactation average model and -0.26 for breeding value estimated with test-day model. The breeding values and pedigree index are based on different scales which explain the negative but favourable correlations. Highest phenotypic correlations were found between the udder type traits fore udder attachment and udder depth (0.56). Pedigree index has a limited use as an aid for selection in a nucleus herd. Cows with the same father and grandfather receive the same pedigree index and their own results are not taken into account. It is better to select and rank a cow of a nucleus herd based on breeding values estimated with an animal model.