Qualitative Behavioural Assessment as a welfare indicator for farmed Atlantic salmon (Salmo salar) in response to a stressful challenge

• Published in: Frontiers in Veterinary Science
• Main Authors: Timothy Robert Wiese, Sonia Rey Planellas, Monica Betancor, Marie Haskell, Susan Jarvis, Andrew Davie, Francoise Wemelsfelder, James F. Turnbull
• Format: Article in Journal/Newspaper
• Language: English
• Published: Frontiers Media S.A. 2023
• Subjects: emotional state; aquaculture; positive welfare; behavioural analysis; qualitative behaviour assessment; Veterinary medicine; SF600-1100; Atlantic salmon; Salmo salar
• Online Access: https://doi.org/10.3389/fvets.2023.1260090; https://doaj.org/article/d4c6f21564df47b89e9a02e21ec541b6

Animal welfare assessments have struggled to investigate the emotional states of animals while focusing solely on available empirical evidence. Qualitative Behavioural Assessment (QBA) may provide insights into an animal’s subjective experiences without compromising scientific rigor. Rather than assessing explicit, physical behaviours (i.e., what animals are doing, such as swimming or feeding), QBA describes and quantifies the overall expressive manner in which animals execute those behaviours (i.e., how relaxed or agitated they appear). While QBA has been successfully applied to scientific welfare assessments in a variety of species, its application within aquaculture remains largely unexplored. This study aimed to assess QBA’s effectiveness in capturing changes in the emotional behaviour of Atlantic salmon following exposure to a stressful challenge. Nine tanks of juvenile Atlantic salmon were video-recorded every morning for 15 min over a 7-day period, in the middle of which a stressful challenge (intrusive sampling) was conducted on the salmon. The resultant 1-min, 63 video clips were then semi-randomised to avoid predictability and treatment bias for QBA scorers. Twelve salmon-industry professionals generated a list of 16 qualitative descriptors (e.g., relaxed, agitated, stressed) after viewing unrelated video-recordings depicting varying expressive characteristics of salmon in different contexts. A different group of 5 observers, with varied experience of salmon farming, subsequently scored the 16 descriptors for each clip using a Visual Analogue Scale (VAS). Principal Components Analysis (correlation matrix, no rotation) was used to identify perceived patterns of expressive characteristics across the video-clips, which revealed 4 dimensions explaining 74.5% of the variation between clips. PC1, ranging from ‘relaxed/content/positive active’ to ‘unsettled/stressed/spooked/skittish’ explained the highest percentage of variation (37%). QBA scores for video-clips on PC1, PC2, and PC4 achieved good inter- and ...