Review: How do we measure welfare on farm? Indicators, proxies and gold standards

Farm animal welfare is assessed for compliance with legislation and labelling schemes, to evaluate the impacts of management change and for animal welfare research. Here, we ask: how can we assess how the animal experiences its life on farm? Animal welfare is what the animal experiences and spans from very negative to very positive emotional states. Welfare is influenced by nutrition, the environment, health and the behavioural interactions that animals have, integrated through the impacts that these have on the mental state of the animal. Whilst we can often readily measure and describe the conditions under which the animal is housed on farm (inputs), and the physical state of the animal within these conditions, these are not sufficient to tell us what the animal is experiencing. The Gold Standard for animal welfare, therefore, would be a measure that tells us something about the mental state of the animal. Currently, this is not considered a directly measurable entity, and it may remain a hypothetical rather than a realistic construct in welfare assessment. Thus, welfare assessment relies on a series of measurable outputs, or indicators, that serve as proxies for the welfare state. Emotional state and welfare are then inferred from these measures. Animal-based measures are considered the most relevant to understanding animal experience, although these require rigorous assessment of the validity and reliability of measurement. Many animal-based measures rely on the assessment of behaviour. These may be ongoing changes in response, such as approach or avoidance, facial expression, Qualitative Behaviour Assessment and circadian rhythms or complex patterns of behaviour, such as preferences, motivational priorities, or perceptions, which allow deeper insight into the likely mental state of the animal. Many measures are only suitable for research purposes at present and can be challenging to assess on farm, but some may be amenable to the use of sensors or automated methods in the future. These have technological and ethical challenges to overcome and require an understanding of what is being measured and how this relates to the animal's mental state but may offer opportunities for continual assessment of animal behaviour and welfare in the future. Currently, however, theoretical understanding of animal welfare, particularly positive welfare and cumulative ‘Quality of Life’, is not well-captured in existing on−farm welfare assessment. We conclude that there is a need for a more comprehensive and integrated approach to the development of methods that can truly address the animal's experience on farm.