How Bright and How Nasty: The Economics of Variable Aposematic Traits

Abstract

Although the theory about the evolution and maintenance of aposematism predicts distinct and uniform warning signals, variation in both signal and defence is common across many taxa, and some species also show correlations between signal and defence. One hypothesis to explain these correlations is honest signalling of defence in prey through the framework of resource competition enforcing a tradeoff. Competition for antioxidant molecules that have dual functions as pigments and in protecting against oxidative stress from toxin sequestration or production have been suggested as a specific candidate resource limitation that could explain warning signal honesty. In Chapter 2 I report an experiment using the large milkweed bug (Oncopeltus fasciatus) as an aposematic model prey that naturally varies in colour and toxicity. By raising milkweed bugs on diets of controlled toxicity and measuring their defence, signal expression, and oxidative stress, I test the expectations of the resource competition model. I found that milkweed bugs overall did not show a correlation between signal in terms of colouration and their level of chemical defence, but that there was a relationship between signal and glutathione amount, a measure of total antioxidant capacity, in the most toxic bugs. These results suggest a mechanistic link between oxidative stress, warning signals and chemical defences in large milkweed bugs. In Chapter 2 I briefly review the concept of dietary wariness, and in Chapter 4 I apply this in practice, using praying mantids (Hierodula membranacea) as a generalist invertebrate predator. I present them with O. fasciatus raised on one of three diets that differ in chemical defence, one species of nontoxic seeds (Helianthus annus) and two different species of their toxic host plants (genus Asclepias). I tested mantids with nontoxic milkweed bugs and measured their level of neophobia and dietary wariness, then tested them with the Asclepias-raised bugs to measure their avoidance learning. Mantids did not learn to avoid the milkweed bugs, even when possibly facing chronic poisoning from consuming them. My results suggest that avoidance learning of toxic prey in predators is not universal. These experiments further our understanding of variation in aposematic traits in prey by examining two key but less explored hypotheses for why signals might vary: resource competition and dietary wariness.