Genetic variation of milk flow and milkability in first lactation Finnish Ayrshires

In dairy cattle, milk flow is an important functional trait which impacts production. Milk flow can be measured accurately by electronic milking meters (EMM) and robots. It is necessary to understand the implications of the transition from the use of subjective scores (milkability, from very slow - 1 to very fast - 5) to objective measurements (milk flow, kg/min), as well as the genetic (co)variation of the traits and with production and health traits. Records from Finnish Ayrshire primiparous cows were analysed for milkability, milk flow, annual milk yield and somatic cell count (SCC). Milk flow was recorded by Tru-Test (EMM) and the Lely robot milking systems. A total of 64 696 cows were analysed for milkability, 1618 cows for Tru-Test and 2232 cows for Lely. To estimate variance components, heritability, genetic and phenotypic correlations, both single and two-trait animal models were fitted and analysed under REML using the DMU software. Heritability of milkability was 0.25 (standard error 0.01). For milk flow, heritability was 0.41 (0.08) and 0.52 (0.08) for Tru-Test and Lely, respectively. The genetic correlations with milk yield were 0.10 (0.04), 0.43 (0.14) and 0.37 (0.14), and with SCC, 0.50 (0.04), 0.42 (0.17) and 0.35 (0.17) for milkability, Tru-Test and Lely, respectively. Common sires provided a way to find the correlation between milkability and milk flow in the absence of common records; the results were positive but low. In conclusion, selection for milk flow is more efficacious than for milkability demonstrating the influence of quality and volume of recording on estimating heritability and genetic correlation