Abstract
Warning signal variation is ubiquitous but paradoxical: low variability should aid recognition and learning by predators. However, spatial variability in the direction and strength of selection for individual elements of the warning signal may allow phenotypic variation for some components, but not others. Variation in selection may occur if predators only learn particular colour pattern components rather than the entire signal. Here, we used a nudibranch mollusc, Goniobranchus splendidus, which exhibits a conspicuous red spot/white body/yellow rim colour pattern, to test this hypothesis. We first demonstrated that secondary metabolites stored within the nudibranch were unpalatable to a marine organism. Using pattern analysis, we demonstrated that the yellow rim remained invariable within and between populations; however, red spots varied significantly in both colour and pattern. In behavioural experiments, a potential fish predator, Rhinecanthus aculeatus, used the presence of the yellow rims to recognize and avoid warning signals. Yellow rims remained stable in the presence of high genetic divergence among populations. We therefore suggest that how predators learn warning signals may cause stabilizing selection on individual colour pattern elements, and will thus have important implications on the evolution of warning signals.
Highlights
Aposematic visual signals are used by prey to indicate unprofitability and/or toxicity to potential predators
We investigated the hypothesis that warning signal variation can be explained by variation in selection on individual pattern elements
We propose that while limited gene flow can permit variation in colour patterns, the mechanisms behind predator learning may allow for stabilizing selection on more salient pattern elements
Summary
Aposematic visual signals are used by prey to indicate unprofitability and/or toxicity to potential predators. Warning signals are often variable both between and within populations of aposematic prey [1] Such variation might be facilitated through non-adaptive processes such as genetic drift and restricted gene flow [2,3]. An alternative hypothesis is that predators may only learn avoidance of warning signals based on individual signal elements (colour, pattern, or shape) of the aposematic signal, and only these elements are under stabilizing selection. We investigated warning signal variation of individual pattern elements in populations of aposematic prey and examined whether potential predators used a signal-elemental approach when learning avoidance of warning signals. In SE Australia, G. splendidus is characterized by a white mantle with a red-spotted colour pattern, encircled by a conspicuous yellow rim (electronic supplementary material, figure S1a). Examining population structure can help determine if signal divergence is occurring in the presence of high genetic differentiation, or whether genetic isolation may be driving the fixation of phenotypic variation
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
