Quantifying Behaviour of Migratory Fish: Application of Signal Detection Theory (SDT) to Fish Passage Research

Behavioural devices used to attract or repel migrating fish to preferred routes of passage are increasingly employed to enhance the efficiency of mechanical mitigation structures (e.g. screens that divert fish away from turbine intakes), or as an alternative to traditional approaches when their application is deemed unviable.  Fish response to stimuli (signals) encountered, and consequently the efficiency of these behavioural devices, can be highly variable.  To further develop efficient fish passage technology based on influencing fish behaviour, there is a need for a generic framework to 1) better understand fish response to environmental stimuli and causes of variation, and 2) quantify ability to detect the stimuli and respond in a way deemed appropriate from a management perspective.  This paper considers application of signal detection theory (SDT) to fish passage research, using downstream migrating fish, primarily juvenile salmonids, as an appropriate model.  Support is provided in the scientific literature for the two key assumptions that downstream migrants are able to exhibit volitional behaviour in response to stimuli encountered and that these behaviours can be interpreted by an observer to indicate the detection of a signal.  Re-evaluation of recently acquired and published data supports predictions of SDT that decision making and elicitation of behaviour is temporally variable within and between individuals, and that sensitivity (e.g. d’ or A’ or other similar indices), a measure of the ability to detect a signal (hydraulic gradient), increases with signal strength and when multimodal stimuli are presented.  Although caveats and areas of future research are suggested, SDT is likely to provide a useful first base for understanding and quantifying fish behaviour in the development of fish passage technology.