Genome‐wide associations for investigating time‐dependent genetic effects for milk production traits in dairy cattle

• Published in: Animal genetics Vol. 43; no. 4; pp. 375 - 382
• Main Authors: Strucken, E. M, Bortfeldt, R. H, de Koning, D. J, Brockmann, G. A
• Format: Journal Article
• Language: English
• Published: Oxford, UK Blackwell Publishing Ltd 01.08.2012; Wiley Subscription Services, Inc
• Subjects: Animals; Cattle; Cattle - genetics; dairy cattle; Female; Genetic Association Studies - methods; Genetic Markers; Genetics and Breeding; Genetik och förädling; genome-wide association study; Genomes; Genomics - methods; Hogs; Holstein; Holstein Friesian dairy cattle; lactation; Lactation - genetics; lactation curve; lactose; loci; Milk; Milk - metabolism; milk yield; Models, Biological; Phenotype; phenotypic variation; Quality Control; Quantitative Trait Loci; single nucleotide polymorphism; somatic cells; Studies; urea nitrogen; variance; Wilmink curve
• ISSN: 0268-9146; 1365-2052; 1365-2052
• DOI: 10.1111/j.1365-2052.2011.02278.x

Phenotypic variation in milk production traits has been described over the course of a lactation as well as between different parities. The objective of this study was to investigate whether variation in production is affected by different loci across lactations. A genome‐wide association study (GWAS) using a 50‐k SNP chip was conducted in 152 divergent German Holstein Friesian cows to test for association with milk production traits over different lactations. The first four lactations were analysed regarding milk yield, fat, protein, lactose, milk urea nitrogen yield and content as well as somatic cell score. Two approaches were used: (i) Wilmink curve parameters were used to assess the genetic effects over the course of a lactation and (ii) test‐day yield deviations (YD) were used as a normative approach for a GWAS. The significant effects were largest for markers affecting curve parameters for which there was a statistical power <0.8 of detection even in this small design. While significant markers for YDs were detected in this study, the power to detect effects of a similar magnitude was only 0.11, suggesting that many loci may have been missed with this approach in the present design. Furthermore, all significant effects were specific for a single lactation, leading to the conclusion that the variance explained by a certain locus changes from lactation to lactation. We confirm the common evidence that most production traits vary in the degree of persistency after the peak as a result of genetic influence.