Do fish cooperate to inspect predators?

Cooperation is a fundamental process of life, ranging from bacterial interactions to human nations. Cooperation can take various forms, including foraging, reproduction and defence. In the latter, inspection of predators by prey animals presents a striking example of cooperation. Always fleeing can be costly. Hence, prey animals should inspect potential predators to determine whether they pose a threat. While beneficial, predator inspection is inherently risky. By cooperatively approaching with a companion, inspectors can dilute their individual risk of capture, should the predator attack.

This has been well studied in fishes. The fish closest to the predator is at higher risk of capture than their trailing partner. Since the costs can be unequal, fish need to ensure cooperation by only continuing their approach if their partner remains alongside them. Because of those early studies, predator inspection in fish had been presented as a textbook example of reciprocal cooperation. However, those studies have attracted much criticism. It has been argued that apparent cooperative predator inspections can be explained by selfish behaviour, i.e. partners balancing orientation towards the predator and social attraction to their companion. What looks like cooperation, critics suggest, is just by-product mutualism, an incidental effect of shoaling in presence of danger.

Fortunately, recent innovations have produced novel experimental and statistical methods that might resolve this 30-year-old debate. First, to distinguish between reciprocity and mutualism the candidate will quantify the movement of lone fish and pairs when approaching real predators, non-predatory stimuli and a control. Statistical models will allow information flow to be inferred and social effects on one another's movement to be quantified. Contingent movements and lead switching are expected in pairs presented with predators, if cooperation is occurring. Second, animated computer-controlled model conspecifics will be developed in conjunction with high-resolution motion tracking to present fish with cooperating or defecting partners to identify contingent investment in cooperative inspection, the basis of reciprocity. In carrying out both parts, fish from populations with and without predators, and laboratory bred F1 offspring will be compared, allowing the candidate to quantify the influence of genetics and experience.

This project will elucidate an important paradox: why individuals cooperate. Although research has been conducted over decades to explain why cooperation exists, we still do not understand its widespread occurrence and mechanisms enabling stable cooperation, of which one solution might be reciprocity. Demonstrating reciprocity in fish will challenge the widespread belief that reciprocity is limited to humans.

Please list any agreed training requirements:


Nothing specific. The candidate will receive standard training through the University's grad skills programme and through selection of appropriate EastBio courses. This will cover but is not limited to statistical training, project management and delivering effective presentations.

The candidate has already been trained in fish welfare, handling and approved S1K techniques. The candidate is expected to (and has already commenced) demonstrating to and tutoring undergraduates and will receive ongoing training and support.

The candidate will be encouraged to take part in external courses and programmes that are relevant or transferrable as these arise. These may include workshops in quantitative skills, programming or outreach for example.