5. Conclusion and perspectives

The first principle to agree upon is that welfare is to be assessed at the level of the animals, and eventually at the level of the individual animal since it is a personal subjective experience. Designing an optimal environment is a sound approach but is not sufficient to ensure welfare. At the level of the organism adaptation is a complex, multifaceted process in which individual psychobiological characteristics – that result from the interaction of genetic and developmental factors – shape the perception of the environment and coping responses. This process takes place in the context of homeostasis. Since we are not yet able to reach directly the cognitive and emotional state of the pig, we have to rely upon the expression of its psychological state in biological functioning, behavioural expression, production efficiency and eventually pathological outcomes. However, most of the criteria available are not specific to stress and welfare problems, since they are sensitive to many influences related to their role in homeostatic processes, or, in the case of pathology, to the action of specific pathogens. It is therefore critical to confront different approaches for a comprehensive assessment of welfare (Figure 6).

In short term studies, a wide range of measures are available. Probably the most difficult problem here is the interpretation of the data in terms of welfare. At least, we have now enough evidence to avoid conceptual oversimplifications like equating an increase of circulating cortisol levels with stress, or biological stress responses with bad welfare. Taking into account the multivariate nature of the response to environmental challenges helps interpreting the results, such as using several biological indices instead of cortisol levels alone, or confronting biological and behavioural responses. Another open question is the extent of individual vulnerability. It is now well documented that behavioural reactivity in animals and mood disorders in humans, as well as biological stress responses are largely dependent upon genetic and developmental factors and more studies should be done in pigs to explore these processes and understand the mechanisms of individual variability of psychobiological reactions to environmental factors.

In chronic studies, the experimental approach is still useful. Indeed, as compared to acute stress responses, allostatic adjustments of physiological systems of animals subjected to sustained stress has not been explored thoroughly. On the other hand, as stressed by Rushen (2003), the epidemiological approach is underutilized. Indeed experimental approaches of system-to-system comparison do not take into account the huge diversity characterizing the animals’ environment. The examples of ecopathological studies presented earlier clearly show that individual environmental factors can be found to influence pathological outcomes. This approach is however a formidable challenge if we were to monitor a wide range of parameters from the animals, to be related to a description of the environment as exhaustive as possible, including its social, physical and human components (Bracke et al., 2002 a,b). Nevertheless, it appears to be unique to take into account the multiplicity of the factors contributing to animal welfare. In these studies still more than in acute studies, it will be important to take into account individual variability that deserves a more sustained interest.