Currently, no specific animal-based measures (ABMs) protocols are available for geese in commercial meat production systems. Following a critical review of the literature and consultation of experts, seven ABMs, potentially valid and feasible for the on-farm welfare assessment of geese, were identified and then tested in 12 farms in Poland to assess their inter-observer reliability. Two observers conducted the assessment, which was divided into two phases. First, a handling test assessed the human-animal relationship (HAR), and a 100% inter-observer reliability was achieved by the observers when evaluating the attitudes of stockpeople and the reactions of geese to humans. Next, an animal inspection was conducted, and the observers simultaneously and independently visually evaluated 100 randomly selected geese per farm and assessed whether the selected ABMs could be identified. In terms of inter-observer reliability, high correlation coefficients were found for plumage dirtiness (ρ = 0.745; p < 0.01), twisted wings (ρ = 0.890; p < 0.001), and broken/twisted wings (ρ = 0.858; p < 0.001). The results showed that plumage dirtiness, twisted wings, and broken/twisted wings are valid and reliable measures. Further research should address the reliability of ABMs of geese in other types of production systems.
Keywords: animal welfare; animal-based measure; goose; inter-observer reliability.