Social behaviour and learning in archerfish

Abstract

Learning is crucial to making behavioural adjustments and fish are adept at using the behaviour of others to more easily find and discriminate between different sources of food, recognise novel predators and other challenges for life in a dynamic environment. This social learning is a powerful process known to impact the success and survival of all animals, humans and non-human animals alike. Yet we understand little about the processes that underpin social learning, and the taxonomic breadth of various mechanisms through which animals can learn from others. Archerfish represent an interesting species to test social learning as there is some evidence to suggest they may be capable of social learning through imitation, currently considered taxonomically restricted and not shown for fish. They are also, unlike most species used in studies of social learning, not highly social: if proven capable of social learning they would afford opportunities to compare social learning species with different levels of sociality. In this thesis I present the first studies exploring archerfish as candidate species for social learning and explore some of the factors that may affect learning performance in experimental conditions. I found that archerfish are sensitive to the presence of conspecifics: increasing their latency to shoot and frequency of orientation prior to shooting. I also found that archerfish display consistent inter-individual differences in latency to shoot and these differences relate to differences in learning speed and discrimination success. Fish with low latencies to shoot learn to shoot novel rewards faster, but fish with higher latencies to shoot achieve greater discrimination success. Despite this sensitivity to others, there was no obvious effect of social presence on learning rates. Finally, I present the results of an initial study of their social foraging dynamics – groups of fish show consistent foraging orders where specific individuals within the group where individuals could be ranked from most to least likely to perform a foraging behaviour when first presented with a prey. Individuals show consistent differences in proportion of foraging activity across successive foraging opportunities - a form of temporal resource partitioning. These results increase our understanding of the behaviour and, by highlighting factors that can may impact their learning, will hopefully be useful in designing further studies of social learning in this unique fish species.