Growing slowly with more space: effects on ‘positive behaviours’ in broiler chickens

Broiler welfare has increasingly been under the scrutiny of the media and non-governmental organisations such as the European ‘Better Chicken Campaign’. Much of the focus has surrounded welfare issues associated with breeding for rapid growth. However, rearing conditions, including stocking density, have also raised concerns. This study explored the welfare of broilers in 4 treatments representing 4 commercial systems with varying combinations of stocking density and breed. The treatments were: (1) Breed A, 30 kg/m2; (2) Breed B, 30 kg/m2; (3) Breed B, 34 kg/m2; (4) Breed C, 34 kg/m2. Breeds A and B were ‘slower-growing’ breeds and Breed C was a typical ‘fast-growing’ breed. Treatments were pseudo-randomly allocated to 4 houses (similar in layout, size and orientation) over 4 crops on one farm. The day before processing, 100 randomly-selected birds/flock were gait scored using a 6-point system (0=no detectable abnormality; 5=incapable of sustained walking).We also compared the levels at which behaviours considered indicative of positive welfare (i.e. appearing playful, exploratory, comfort-related or safety-related in the context observed) were expressed in the different flocks. We sampled behaviour using ‘behaviour transect’ methodology, whereby successive 15-s scans were made of 57 ‘disturbed’ (D) patches and 54 ‘undisturbed’ (U) patches per house. A D patch was a defined area within a transect through which the assessor had slowly walked 75 s previously, while a U patch was a defined area located two transects across from, and ahead of, the assessor’s current location. The assessor recorded numbers of birds/patch/scan performing the following mutually-exclusive behaviours: running, jumping, wing-flapping, play-fighting, ground-scratching, dustbathing and perching, from which the overall percentage of birds engaged in ‘positive behaviour’/patch/scan was calculated. Behavioural transect sampling was undertaken at 14 and 28 days old and again 2 days before processing. Clear differences in the behaviour and health measures were detected between treatments. There was a difference in the distribution of gait scores between treatments (Kruskal Wallis, Χ2 (3)=368.7, P