opinion, published on 22 September 2009, replaces the earlier version published on 24 February 2009 
Following a request from the European Commission, the AHAW Panel was asked to deliver a Scientific Opinion on the animal welfare aspects of husbandry systems for farmed fish. In addition to the already adopted scientific opinions regarding six different species of farmed fish, a scientific opinion on the general approach to fish welfare and to the concept of sentience in fish was adopted on 29 January 2009.
The scientific opinion focused on the neurobiology and special sense organs in fish, and their capacity to experience pain, fear and distress, expressions of sentience from a fish welfare point of view, taking into consideration the available scientific information. Only a small number of the circa 20,000 teleost fish species, from diverse groups, generally those of economic importance for fisheries or aquaculture, have been studied in any scientific detail. It is therefore important that generalisations across the range of species should not be used without qualification.
It was concluded that the concept of welfare is the same for all the animals, i.e. mammals, birds and fish, used for human food and given protection under the Treaty of Amsterdam. However, fish welfare has not been studied to the same extent as mammals and birds. Whilst similar measures of welfare developed for other animals are often relevant to fish, clearly defined protocols for fish welfare evaluation are lacking.
Due to the complex relationships among the various needs/requirements of farmed fish and their behavioural and physiological consequences, as for all animals it is impossible to find one single measurement or welfare indicator that will cover all possible husbandry systems, farmed species and situations. A range of welfare indicators should be considered when welfare is being evaluated. Indicators of fish welfare should be species-specific, validated, reliable, feasible and auditable.
Different species of fish have evolved highly sophisticated sensory organs to survive in changing and varied environmental conditions. There is scientific evidence to support the assumption that some fish species have brain structures potentially capable of experiencing pain and fear. The balance of evidence indicates that some fish species have the capacity to experience pain. However research and developments in the area of cognition and brain imaging techniques should be carried out in fish to further our knowledge and understanding of pain perception.
Defence and escape behaviours are dependent on cognitive and learning abilities related to fear. Responses of fish, of some species and under certain situations, suggest that they are able to experience fear. Fish possess a suite of adaptive behavioural and physiological responses that have evolved to cope with stressors. Many of these are homologous with those of other vertebrates. Fish show short term adaptive responses which may be important to the maintenance of homeostasis but these do not necessarily imply any harmful consequences. Prolonged exposure to stressors generally leads to maladaptive effects or chronic stress. Chronic stress responses indicative of poor welfare include reduction in immune function, disease resistance, growth and reproduction, eventually death. Cumulative stress responses occurring at different life stages have not been studied.
From studies of sensory systems, brain structure and functionality, pain, fear and distress there is some evidence for the neural components of sentience in some species of fish. Our knowledge and understanding of manifestations of sentience in fish, however, are limited.