Use of eye white and heart rate as dynamic indicators of welfare in Holstein dairy cattle

A. L. Tucker1*, B. Zouaoui1, T. J. DeVries1 and R. Bergeron1


The percentage of visible eye white (%EW) has been recently identified as a novel and dynamic indicator of emotion in Norwegian Red dairy cattle. Fear and frustration have been associated with opening of the eye (more visible eye white), while positive experiences, such as feeding and calf-cow reunions, have been associated with a reduction in eye white. Heart rate (HR) has also been used extensively as an indicator of welfare in cattle with negative experiences tending to increase cardiac output.

The aim of this study was to examine the association between % EW and HR under both a positive rewarding stimulus and a negative fearful stimulus in the most common breed of dairy cattle in North America, the Holstein. Over a 3-d period, 12 multiparous Holstein cows were subjected to both a positive stimulus (provision of food and water) and negative stimulus (startling with umbrella). Heart rate monitors were used to record heart rate at 5-sec intervals for 5 min before and after each treatment while 1 fixed video camera collected behavioural data continuously. A second hand-held video camera (held at a distance of 2 m lateral to the cow) collected eye white data. Still eye white photos were selected for analysis when the cow was facing perpendicular to the hand-held camera and when the cow was neither distracted nor performing any other behaviours (e.g. grooming). Only 1 photo was collected from each 30-sec period and was digitally analyzed. For each stimulus, a general linear mixed model was used to assess the association between % EW and HR; cow and day were included as random effects in the model. Body condition score, weight, parity, stage of lactation and time of day were included as covariates.

Heart rate was positively associated with % EW under both the positive (P=0.003) and negative stimulus (P=0.004). Prior to feeding, mean HR and % EW were 94.3 ± 8.6 bpm and 30.2 ±1.8%, respectively; these decreased to 91.9 ±5.9 bpm and 27.9 ± 1.2% after feeding (P=0.013). Before opening the umbrella, mean HR and % EW were 74.3 ± 2.2 bpm and 26.5 ± 0.87%, respectively; these increased to 76.5 ± 4.9 and 30.7 ± 1.8% (P=0.099) after opening the umbrella. Closure of the umbrella resulted in a decrease in HR and % EW (P=0.012) to 75.8 ± 2.5 bmp and 26.3 ±1.3%.

These results indicate that EW responses in Holstein cattle are consistent with those reported for Norwegian Red cattle and that % EW is strongly associated with HR under both positive and negative conditions. We suggest that % EW be used as a non-invasive measure for assessing affective state in Holstein dairy cattle.


Proceedings of the 5th International Conference on the Assessment of Animal Welfare at Farm and Group Level. 8-11 August 2011. Guelph, ON.

Author Locations and Affiliations

(1) University of Guelph, Department of Animal & Poultry Science, Guelph, ON, N1G 2W1
(2) National Agronomic Institute of Tunisia, Department of Fish and Animal and Agri-Food Technologies, Tunis, Tunisia
* Corresponding author, E-mail

Posted May 2013