Colour, size and movement as visual subcomponents in multimodal communication by the frog Allobates femoralis

Multimodal signals convey information that can be detected by receivers using two or more sensory modalities. The information transmitted through secondary channels may be mostly useless, redundant, complementary, or essential. Although each option implies different evolutionary scenarios, the relative role of signal subcomponents in a secondary channel has rarely been tested. Anuran amphibians are known for their extensive use of auditory signals, but growing evidence suggests their common use of multimodal signals where the visual component represents the secondary channel. Territorial males of the nontoxic frog Allobates femoralis use bimodal (acoustic + visual) signals to recognize conspecific intruders. We used playback experiments to test the relative importance of three visual subcomponents (body coloration, body size and body movement) in a multimodal signal to elicit agonistic reactions in males. The probability of attack was highly related to the movement and size of a dummy frog but very poorly related to dummyʹs coloration. The results suggest that any body movement, and not only vocal sac movement, is a nonredundant component of a bimodal signal that probably modulates male aggression. Against our expectations, males readily attacked dummy frogs painted to mimic the co-occurring frog Ameerega trivittata as well as models lacking some of their own patternʹs subcomponents (white lines and yellow patches). Colourful frogs did not use body coloration as a visual cue for conspecific recognition. Both patterns recall the importance of carefully manipulating the availability of visual information in interpreting the origin and evolution of multimodal signals.