Document type : Article from the 25th Rencontres Recherche Ruminants
Authors: M. Berthelot, M.M.Mialon, A. Aupiais, A. Boissy, C. Aubert, C. Baudry, V. Michel
Preview: The development of welfare indicators adapted to the different stages of development of animals is essential to ensure their welfare throughout their life. This work focuses on the study of inter-observer reproducibility, intra-observer repeatability and the stability of welfare indicators around the time of breeding of goat kids between seven and twelve months.
Individual measurements (body condition, hoof condition, lameness, reaction to an approaching human, etc.) were taken from 451 kids on 10 farms. Batch level measurements from locations on these ten farms were also taken: waiting at the watering and feeding area, synchronisation of behaviour between animals, reactions to a stationary human, to a new object and to a sudden event, the presence of stereotypies, the presence of prostrate and heat stressed animals, and the emotional state of the animals (ABQ). All measurements were recorded simultaneously by two observers on two consecutive days. Inter-observer reproducibility, intra-observer repeatability and stability of measurements were assessed using Kappa coefficients (K) for qualitative variables and intra-class correlation coefficients (ICC) for quantitative data. For the individual measurements, reproducibility and repeatability were good for most of the measurements (K>0.61), except for anal region cleanliness, hair condition, body condition and discharge. For the measurements taken from the batches, the reproducibility was good to excellent (CIC>0,75), except for the ABQ where it was average. The stability of the measurements from one day to the next was good for animals waiting during feeding and reactions to an approaching human, whether in the batch or at the feed fence, and to a new object. The stability of reactions to a stationary human and to a sudden event seems to be influenced by the familiarity of the goats with these tests. These results will allow the selection of reproducible, repeatable and stable measures that can be included in a general assessment tool for goat welfare.
