Warning and Deception: Chemical, Behavioral, and Phylogenetic Studies of Aposematic Coloration and Mimicry

Abstract

The study of aposematic coloration and mimicry has a long and distinguished history, and has stimulated scientific inquiry in areas as diverse as chemistry, evolution, ecology, and behavior. Yet, many questions regarding signal function and ecological dynamics remain unknown. This dissertation attempts to address some of these questions about how a visual warning signal functions and how the environment changes its efficacy. First, I evaluated the role of luminance contrast in aposematic signaling using milkweed bugs as model prey and Chinese mantids as model predators. Predators learned to avoid unpalatable prey sooner and remembered to avoid unpalatable prey for longer when the prey had higher luminance contrast with the background. These results help define what makes a visual signal conspicuous and designate the importance of high luminance contrast in the efficacy of a warning color signal. Another important characteristic of warning coloration is the reason for the advertisement. I was able to identify and quantify the toxic compounds in both the host plant and the viceroy butterfly, a putative aposematic insect. These results provide a chemical mechanism for previous research that demonstrated that the viceroy was unpalatable to avian predators. Next, I was able to test the role of geographic variation in host plant and viceroy chemical defense and how that variation compared with the local abundance of a mimicry co-model of the viceroy, the queen butterfly. The results indicated the viceroy was more chemically defended and more unpalatable in locations where the queen was at low abundances. This result suggests that mimicry evolves in a geographic mosaic of co-evolution. Finally, I used molecular phylogenetic approaches to reconstruct and test the evolution of mimicry in the North American admiral butterflies (Limenitis: Nymphalidae). One species, L. arthemis, evolved the black, pipevine swallowtail mimetic form but later reverted to the white-banded ancestral form. This character reversion is strongly correlated with the geographic absence of the model species and its host plant, not the mimics host plant distribution. These results support the idea that loss of model in a geographic area is not an evolutionary stopping point for a Batesian mimic.