Ornamental evolution in Trinidadian guppies (Poecilia reticulata): insights from sensory processing-based analyses of entire colour patterns

摘要

The evolution of exaggerated sexual ornamentation is classically thought to proceed as a compromise between opposing vectors of sexual and natural selection. In colour-based ornamentation, as exhibited by guppies (Poecilia reticulata), heightened trait expression may be beneficial in promoting attractiveness, but costly in terms of predation. Opportunities to reconcile this compromise will exist if there are differences between conspecifics and predators in their sensory systems; in such situations guppies should evolve to exploit the ways in which their ornamentation would appear maximally conspicuous to conspecifics. In the present study, we addressed this hypothesis via a study of geographic variation employing the most sophisticated colour analysis yet attempted for Trinidadian guppies. We made two paired contrasts, one between two Aripo populations that vary in the presence of the potential predator Aequidens pulcher, and another between Quare and Marianne populations that vary in exposure to a predatory prawn, Macrobrachium crenulatum. We predicted that, if ornamentation is constrained by the presence of either predator, then guppy conspicuousness should change most markedly across each of the two paired populations as viewed by that predator. Although disparity analysis of entire colour patterns indicated significant differences in both contrasts, this prediction was most clearly supported for the Marianne/Quare contrast. Marianne fish, which co-exist with prawns, exhibited larger black spots coupled with less extensive, less bright flank iridescence. The brightness reductions are notable because, as the only potential guppy predator with a dedicated ultraviolet (UV) photoreceptor, prawns may detect passing male guppies via their UV-bright blues, violets and 'UV/oranges'. We discuss our findings in light of the additional insights that might be obtained by combining spectral assessments and visual modeling with more traditional methods of colour pattern appraisal.