As aquaculture expands and scientific evidence for fish sentience grows, developing reliable, practical welfare monitoring methods for farmed fish has become a research priority. The challenge is to translate welfare science into on-farm tools that enable meaningful assessment without disrupting production systems.
Several factors make fish welfare monitoring particularly challenging compared to terrestrial livestock: fish are maintained in three-dimensional water environments making individual observation difficult; large numbers make individual assessment impractical; fish physiology differs fundamentally from mammals, making welfare indicator transfer from terrestrial species inappropriate; validated fish welfare indicators are fewer than for most terrestrial livestock species; and the diversity of farmed fish species means species-specific indicator development is needed.
Operational Welfare Indicators (OWIs) are welfare-relevant measures that can be assessed practically in commercial aquaculture settings. Examples for Atlantic salmon include: eye index (opaque or damaged eyes indicating poor welfare), fin damage prevalence (fin erosion from crowding, environmental factors), external injuries and wounds, snout damage (from net interactions), opercular abnormalities, and behaviour assessment (surface breathing indicating hypoxia, schooling disruption patterns). OWIs provide a practical welfare snapshot without requiring individual fish assessment.
Collective fish behaviour provides welfare information. Normal schooling patterns, feeding behaviour (response to feed presentation, feed conversion efficiency), vertical distribution (surface vs. mid-water vs. bottom preference), and spontaneous activity levels reflect welfare status. Stressed fish show: surface crowding (seeking oxygen), abnormal swimming patterns (spiral swimming, belly-up behaviour indicating disease or toxicity), reduced feeding response, and disrupted schooling. Computer vision analysis of fish movement is enabling automated, continuous behaviour monitoring in research settings.
Physiological indicators measured on sampled fish include: plasma cortisol (primary stress indicator — elevated in acute stress but complex to interpret in chronic stress due to adaptation), plasma glucose (elevated in acute stress), haematocrit (anaemia indicator — important for lice infestation welfare assessment), white blood cell counts (immune function assessment), and opercular beat rate (a non-invasive proxy for stress when measured by video analysis). Blood sampling itself causes stress, so indicators must be interpreted in context.
Environmental parameters are not direct animal welfare indicators but serve as welfare-relevant proxies: dissolved oxygen (welfare critical — fish experience distress at low DO levels), ammonia and CO2 (toxic at elevated concentrations), water temperature (affects metabolic rate, disease risk, and welfare), and stocking density (indirectly affects behaviour, injury rates, and water quality). Continuous monitoring of these parameters via farm management systems provides early warning of welfare-threatening environmental deterioration.
Welfare monitoring tools developed in research need translation to practical farm tools. Simple scoring sheets for OWIs, mobile phone apps for recording welfare indicators, integration with farm management software, and training programmes for farm staff are needed to operationalise welfare science. The Aquaculture Stewardship Council, RSPCA, and national aquaculture industry bodies are developing welfare indicator protocols and training resources. Incentivising welfare monitoring through certification and market premiums creates the commercial environment for adoption.