Species discrimination in guppies: Learned responses to visual cues

The possibility that animals can discriminate conspecifics and heterospecifics by integrating multiple-cue information obtained by a single sensory modality was tested using guppies, Poecilia reticulata. The influence of familiarity on guppy discrimination was examined through binary-choice experiments in which juvenile guppies were allowed to associate with familiar or unfamiliar siblings, unrelated conspecifics or heterospecifics (swordtails, Xiphothurus helleri). Stimulus and subject (test) individuals were separated by partitions made of clear Perspex or plastic mesh, and the results of trials using the two partition types were compared to provide insight into the sensory modalities involved in recognition. Guppies associated more readily with familiar than unfamiliar conspecifics. The trials with the Perspex partition revealed that visual cues were involved in conspecific discrimination. Guppies were also visually attracted to familiar swordtails, and to unfamiliar swordtails as long as the guppies were originally reared with swordtails, showing that the formation of a recognition template in early life is flexible and not controlled by a closed genetic programme. However, there was evidence that attractive long-range swordtail cues were overridden by those permitting more detailed species discrimination at close range. In the trials with the mesh partition, attraction to siblings was apparent only at close range, suggesting that the mesh impaired the visual discrimination abilities evident from the Perspex trials. Heterospecific attraction to swordtails increased when the mesh partition was used, possibly for the same reason. It is likely that guppies integrate responses to distance-dependent visual (and possibly mechanical) stimuli to permit recognition of individuals, kin members or species. With regard to discrimination studies on aquatic vertebrates, the potential advantages to be obtained from the use of complementary behavioural assay techniques, and from analyses that take into account inter-brood variation and the length of the experimental acclimation period, are stressed. (C) 1996 The Association for the Study of Animal Behaviour