Archerfish as a model for sophisticated social learning

Supervisors: Luke Rendell, Keith Sillar, Mike Webster

Project Description:

Social learning is defined as ‘learning that is facilitated by observation of, or interaction with, another individual or its products’, allowing animals to gather information and make adaptive choices [1]. It is a vibrant area of current research attracting interest from biologists, psychologists and anthropologists, not least because it is considered a crucial element in both human social development and human evolution. Social learning has probably evolved several times in animals, with several mechanisms having been identified [1].  While recent studies indicate fish are capable of some forms of social learning, no fish species has been shown to be capable of the most cognitively sophisticated form of social learning: imitation. Imitation is the learning of novel motor patterns from observed behaviour, or using a specific action in a novel circumstance [1].

Recent research has revealed surprisingly sophisticated cognitive capabilities in archerfish (Toxotes jaculatrix), including their ability to compensate for underwater distortion and accurately estimate size and distance of targets, learn to shoot moving targets [2], and make rapid context-dependent decisions on target selection. However no study has focused specifically on social learning in archerfish, although recent work in our lab has shown they are highly aware of having an audience [3].

The primary aim of this project will be to reveal the mechanisms archerfish use to learn, by observation of others, how to hit moving targets. Given that archerfish are capable of learning to hit moving targets by observation, they may also be able to use other forms of social information, such as foraging success rates of other conspecifics. The second aim of the study is therefore to understand which kinds of social information archerfish can or cannot use. This represents a hypothesis that the different domains of social learning are in fact linked in evolution – if archerfish do not use social information in this way, it suggests that this kind of learning can be encapsulated and context-specific rather than domain general, whereas if they do, it suggests that social learning and social information use are more general abilities.

This project will be run in as part of an ongoing collaboration with Prof Stefan Schuster, of Bayreuth University Germany, and we plan for staff and student exchanges during the course of the project. It will use an existing archerfish captive stock held at the University of St Andrews, where all the required equipment is available, but will likely also involve experiments run in Bayreuth.

Research in this area has been somewhat restricted by the lack of an easily manipulable laboratory model for studying the neurobiology of social learning. Archerfish have however been well studied from a visual neuroscience perspective because of their ability to shoot down prey using visual cues that cross the air-water interface, so this project could be a first step to establishing a new research programme on the visual neuroscience of imitation.


1. Hoppitt  & Laland (2013) Social Learning: An Introduction to Mechanisms, Methods, and Models. Princeton University Press.

2. Schuster, et al (2006) Animal Cognition: How Archer Fish Learn to Down Rapidly Moving Targets. Current Biology, 16: 378-383.

3. Jones et al. (2018) Presence of an audience and consistent interindividual differences affect archerfish shooting behaviour. Animal Behaviour 141: 95-103

To apply for this project, please go to this link.