Animal welfare in broiler chickens

Abstract

The objective of this study was to assess broiler chickens welfare using transect and individual
assessment methodologies, evaluating the robustness of the latter as an effective tool for welfare evaluation
on poultry farms. Eight closed, tunnel-ventilated houses were studied, four fitted with blackout curtains and
four with translucent curtains. Along each transect, the number of immobile, lame, dirty, sick, dying, or dead
chickens was recorded. In the individual assessment, variables such as body weight, presence of footpad
lesions, tarsal burns, feather cleanliness, and gait were measured at each sampling point. Hierarchical
clustering analyses were performed using three aggregation methods: Complete, Average, and Ward.D,
allowing grouping of farms based on similarity in these variables. Farm-level analysis produced stable
clusters regardless of methodology. Dendrograms from the transect methodology were consistent across
all three clustering methods, while individual assessment also yielded stable clustering independent of the
clustering approach used. The findings suggest that both methodologies generate consistent and robust
patterns in the data, supporting their validity and applicability. Although minor differences between methods
were observed, their results were sufficiently congruent to recommend their combined use in future studies,
expanding analytical capabilities in diverse poultry welfare assessments.