Automated tracking of individual activity of broiler chickens

There is a growing interest in quantifying individual behaviour of group-housed animalsand its relation to individual performance. Broiler chickens are an example of a livestockspecies for which individual data can be valuable. Broiler breeding goals focus on efficientgrowth and reproduction, as well as welfare indicators for balanced genetic improvements.Recording of broiler behaviour can provide insight into welfare indicators, such as activity orgeneral leg health. However, monitoring individual behaviour in group-housed animals is achallenge. Often, video analyses are used, but these are time-consuming and prone to humanerror. Therefore, automated systems for monitoring individual animals are desired. Here, westudied whether individual broiler activity could be tracked using an ultra-wideband (UWB)system. Birds were fitted with UWB tags that sent out signals to four beacons. The location ofthe birds was determined using triangulation of the signal, allowing calculation of distancesmoved over time. Distances moved according to the UWB system were compared to thosefound on video for twelve birds. A moderately strong correlation between the UWB systemand video tracking was found (Repeated measures correlation, r=0.71 (95%-CI: 0.64-0.77),df=209, P<0.001). Furthermore, the UWB system was used for assessing individual levels ofactivity. In total, 137 birds from different genetic crosses were tracked near-continuously forseventeen consecutive days, starting on day 16 of life, and their weight was determined at thestart and end of this 17 d period. Data were analysed using an LME-model in R. First analysesshowed that activity, measured as the average distance moved, decreased over the seventeendays in all genetic crosses (F(1,127.00)=301.47, P<0.001). Furthermore, in all genetic crosses,birds with a lower weight at the start of the trial were on average more active (F(1,125.14)=9.16,P<0.01). Overall, the UWB system appears well-suited for activity monitoring in broilers andthe longitudinal information on individual differences in activity can potentially be used tomonitor health, welfare and performance at the individual level. Unfortunately, the UWBtags are too large and heavy for day-old chicks to wear and can only be implemented later inlife. Therefore, current work is focussing on the implementation of a passive radio frequencyidentification (RFID) system to track individual broiler activity, using smaller, lightweight tagsthat can be attached to the broilers’ legs at day-old. This system may be able to track individualactivity of broiler chickens throughout the entire life.