Repeatability of flatfish reflex impairment assessments based on video recordings

Area of Science:

• Marine Biology
• Fisheries Science
• Animal Welfare

Background:

• Assessing the vitality of discarded non-target fisheries catch is crucial for conservation.
• Reflex impairment and injury measures are increasingly used as predictors of aquatic organism survival.
• Rater subjectivity and expectation bias can potentially influence the reliability of these vitality assessments.

Purpose of the Study:

• To evaluate the robustness of reflex impairment as a survival predictor against rater subjectivity.
• To test inter- and intra-rater repeatability in assessing fish vitality using reflex impairment.
• To investigate the impact of expectation bias on the reliability of vitality assessments.

Main Methods:

• Multiple raters with diverse experience assessed reflex impairment in beam-trawled common sole (Solea solea) from video clips.
• Inter- and intra-rater repeatability were measured using various assessment conditions, including duplicated clips and indicated air exposure times.
• Expectation bias was evaluated by comparing reflex ratings under true and exaggerated air exposure durations.

Main Results:

• High inter- and intra-rater repeatability (intra-class correlation coefficients of 74%) were observed for reflex impairment assessments.
• Neither background type nor expectation bias significantly affected the repeatability of the vitality assessments.
• Scoring consistency varied for certain reflexes based on rater experience, suggesting subjectivity in some measures.

Conclusions:

• Reflex impairment is a robust metric for predicting fish survival, reliable across diverse raters and unaffected by expectation bias.
• Ample rater training is essential for consistent and reliable vitality assessments, with no prior experience necessarily required.
• The subjective nature of certain reflex assessments warrants careful consideration and evaluation before implementation in vitality studies.