Kimberly Vann, Arizona Sate University (Dr. R. Rutowski)
Project title: The role of iridescent coloration in te warning signal of Battus philenor, the Pipevine Swallowtail Butterfly

Warning signals warn predators of prey distastefulness, and increase the speed with which predators learn to associate color pattern with distastefulness.  Most research on warning colors has examined color components that are diffusely reflected oranges, blacks, and reds1.  In contrast, I am investigating a warning signal that has an iridescent (highly directional) and blue component as part of my Ph.D. research.  The signal is the wing coloration of Battus philenor, the Pipevine Swallowtail Butterfly.  In this species, larvae ingest and sequester aristolochic acids, which make them highly unpalatable to predators2,3.  The color pattern on the ventral hindwing features an iridescent blue field spotted with orange.  Birds learn to avoid this butterfly based on the overall ventral hindwing coloration4.  The color patterns of butterfly species that mimic B. philenor include the blue iridescence3, and this strongly suggests that the iridescent blue is a salient part of their warning signal.  Also, the variation in the ventral hindwing iridescence of B. philenor suggests it is naturally rather than sexually selected.5
A critical first step in my research plan is to test whether this blue iridescence is a key feature of the warning signal that birds learn and attend to.  My experiments will feature as a potential predator Curve-billed Thrashers (Toxostoma curvirostre), an omnivorous bird that has been locally observed grabbing and investigating B. philenor as potential prey.  I will first present freshly-killed B. philenor adults to caged birds (n = 26) until they learn to avoid the butterflies, if they have not learned to do so prior to capture.  Then, I will alter the ventral wing patterns using a black marker to remove specific color pattern elements.  I will present these altered butterflies, and record the birds’ reaction.  The order of color pattern eliminations will be randomized among the birds and presentations of B. philenor will be randomly alternated with a palatable butterfly as a control for state of hunger.  I expect that some specific manipulations will result in butterflies not being recognized as distasteful and attacked.  An increase in attack rate when only the iridescent blue is removed will indicate that this component plays a critical role in recognizing this distasteful butterfly.  On the other hand, an increase in attack when only the iridescence remains will indicate the blue is not used in the learning process.
Once this is determined, I will explore other compelling questions.  If the blue iridescence is what birds learn, I will ask what features make it effective. Candidate features include the blue hue, the directionality of the iridescence, or the contrast with other pattern elements.  In contrast, if the blue iridescence is not what the birds learn, its presence on mimics still suggest that it is an anti-predator signal.  It could be a startle or dazzle signal, an attention-alerting signal, or an amplifier signal.  Therefore, either outcome of the proposed experiment will open up lines of investigation that will contribute to our understanding of how color signal properties shape their value as part of a warning signal.