PLoS ONE (Jan 2013)

Can convict Cichlids (Amatitlania siquia) socially learn the degree of predation risk associated with novel visual cues in their environment?

  • Patrick M Barks,
  • Jean-Guy J Godin

DOI
https://doi.org/10.1371/journal.pone.0075858
Journal volume & issue
Vol. 8, no. 9
p. e75858

Abstract

Read online

For many animals, the ability to distinguish cues indicative of predation risk from cues unrelated to predation risk is not entirely innate, but rather is learned and improved with experience. Two pathways to such learning are possible. First, an animal could initially express antipredator behaviour toward a wide range of cues and subsequently learn which of those cues are non-threatening. Alternatively, it could initially express no antipredator behaviour toward a wide range of cues and subsequently learn which of them are threatening. While the learned recognition of threatening cues may occur either through personal interaction with a cue (asocial learning) or through observation of the behaviour of social companions toward a cue (social learning), the learned recognition of non-threatening cues seems to occur exclusively through habituation, a form of asocial learning. Here, we tested whether convict cichlid fish (Amatitlaniasiquia) can socially learn to recognize visual cues in their environment as either threatening or non-threatening. We exposed juvenile convict cichlids simultaneously to a novel visual cue and one of three (visual) social cues: a social cue indicative of non-risk (the sight of conspecifics that had previously been habituated to the novel cue), a social cue indicative of predation risk (the sight of conspecifics trained to fear the novel cue), or a control treatment with no social cue. The subsequent response of focal fish, when presented with the novel cue alone, was not influenced by the social cue that they had previously witnessed. We therefore did not find evidence that convict cichlids in our study could use social learning to recognize novel visual cues as either threatening or non-threatening. We consider alternative explanations for our findings.