Use of accelerometer data for genetic evaluation in dairy cattle

Abstract

Detection of estrus plays an important role in modern reproductive management of dairy cattle. Activity monitoring systems incorporating accelerometer devices are common on-farm tools for the detection of cows in heat and thus for timing AI. The use of these accelerometer data for genetic improvement of the ability of a cow to show estrus requires the definition of meaningful traits with sufficient genetic variance. In the present study, three different activity traits were defined and analysed based on activity data of 1.171 cows. Trait A1 is the general level of the activity of a cow calculated by the simple mean of a cows daily activity value over a time period of at least 100 days. A1 has a low heritability (0.03 – 0.12) and activity levels during the estrus period and the non-estrus period are highly correlated. Trait S1 describes the intradaily deviation from the baseline and is in close correlation to A1 (r_g=0.67). However, the estimated heritability for S1 is remarkably higher compared to A1, and S1 during the estrus period and S1 during the non-estrus period seem to be different traits. Thus, the difference between these two periods was analyzed separately and resulted in heritability estimates between 0.10 and 0.15. These first results for the inclusion of activity data in genetic evaluations are promising but further research is necessary in terms of alternative trait descriptions, sophisticated statistical models, and genetic relationships between activity traits and traits of milk performance and behavior.