Onderwerp scriptie

Inter-observer reliability of behavioural indicators of animal welfare in the Welfare Quality assessment protocol for dairy cattle - Merel Schouten

The Welfare Quality protocol is a protocol to assess animal welfare on dairy farms, using animal-based indicators. The aim of this study was to assess the inter-observer reliability of the behavioural indicators of the Welfare Quality assessment protocol by comparing data of two observers.

The Welfare Quality protocol is a protocol to assess animal welfare on dairy farms, using animal-based indicators. Of the animal-based indicators the inter-observer reliability of behavioural observations were not yet evaluated. The aim of this study was to assess the inter-observer reliability of the behavioural indicators of the Welfare Quality assessment protocol by comparing data of two observers. Two trained observers visited eight Dutch dairy farms three times and conducted the behavioural observations at the same time on the same group of animals. The observed behaviours of the two observers were compared to assess the inter-observer reliability per behaviour. Also a correlation test was done on all behaviours. All the behaviours of the Herd Scan, Segment Scan, Continuous Observations and Time Lying Down were statistically analysed.

The behaviours of the Herd Scan Standing, Standing in Cubicles, and Feeding did not differ between the two observers. However, Lying did differ. Of the Segment Scan, the behaviours Standing, Feeding, Lying, Standing In Cubicles, did not differ between the observers. The behaviours Lying Out and Lying Not Seen were in the Start Scan different, where only the behaviour Lying Not Seen in the End Scan was different between the two observers. Of the Continuous Observations, the behaviours, Chasing, Sneezing, Tongue Rolling, Chasing Up did not differ between the two observers. The behaviours Head Butts, Fighting, Displacement, Social Licking, Coughing and Rubbing differed between the two observers. The behaviours of Time Lying Down; Time Lying Down and Collision with Housing Equipment During Lying Down; differed between the two observers. A strong correlation was found between the two observers for all behaviours, except for Sneezing, Tongue Rolling and Collision with Housing Equipment during Lying Down.

The behavioural observations which differ between the two observers do not seem to be reliable tools useable for the Welfare Quality protocol and can give a skewed reflection of the actual animal welfare on a dairy farm. So it is recommended to reconsider the behaviours, the type of observation in the protocol or the training of the observers to improve the inter-observer reliability.


Student: MC Schouten

Supervisors: dr ir M de Vries, dr ir EAM Bokkers

36 Ects