Open Collections

Replication Data for: Automated, longitudinal measures of drinking behaviour provide insights into the social hierarchy in dairy cows

Sheng, Kehan Sky

Water is an essential resource for dairy cattle, and in some circumstances cattle will compete with one another to gain access to water. Here we applied a Bayesian-based, Elo-rating method to assess the winning probabilities of 87 cows housed in a dynamic group and compared the resulting social hierarchies based on their steepness. We identified a hierarchy at the drinker with a steepness of 0.55±0.02 whereas the hierarchy detected at the feeder during the same time period was less steep (0.45±0.02), indicating smaller average differences between the winning probabilities of cows when competing for feed compared to competing for water. Individual cows’ winning probabilities at the feeder and drinker were moderately correlated (rs=0.55, P<0.001). However, cows at both the lower and upper ends of the hierarchy demonstrated a consistent alignment. We compared drinker hierarchies between periods with THI above and below 72 and found similar steepness (0.54±0.03 and 0.56±0.03 respectively) and the individual winning probabilities of cows were highly correlated between hot and normal periods (rs=0.87, P<0.001). Individual drinking behaviour was also associated with the drinker hierarchy, cows with higher winning probability had lower average daily visit frequency (hot: rs=-0.40, P<0.01, normal: rs=-0.33, P<0.01) and higher average daily water intake (hot: rs=0.38, P<0.01, normal: rs=0.37, P<0.01). We also found evidence that cows differ in when they drink, depending on their winning probability; less successful cows shifted their drinking times to before or after the visit peak after milking. Automatically identifying cows with consistently high or low winning probabilities using drinkers may inform grouping decisions and water provision on farms.

Dataset

University of British Columbia Library

10.14288/1.0439803

https://doi.org/10.5683/SP3/U8TXOC

Dataverse