Insights into butterfly ecology and evolution

This dissertation examined features of the butterfly visual system. The first part used opsin genes, which code for the visual pigment proteins, in reconstruction of butterfly phylogenies. The second part examined butterfly behavioral responses to flower color visual cues. The first goal was to examine the effect of incorporating duplicate opsin gene copies on tree reconstruction and divergence time estimation. Sequences from 5 genes, including 3 opsins, were obtained from 27 taxa. Regardless of the phylogenetic reconstruction method, combined data sets analyses using either slower or faster evolving copies of duplicate genes, as well as individual analyses of blue and long-wavelength opsin genes, which are present in multiple copies in some lineages, resulted in a single topology in agreement with our current understanding of family relationships. Two methods resulted in similar divergence time estimates regardless of whether faster or slower evolving copies were used. Family-level results were congruent with other recent estimates, indicating an age of at least 150 MY for the first familial divergence. These results are consistent with overlapping timeframes for the diversification of butterfly families and angiosperms and suggest the use of duplicate gene copies for phylogenetic reconstruction and divergence time estimation.;The second goal was to explore the role of color in flower visiting behavior. This study represented the first field-based attempt with butterflies to phenotypically manipulate flower color and decouple its effect from that of other traits. In order to assess if butterflies possess spontaneous color preferences and if they behave as constant flower visitors, all visits to artificial arrays of Asteraceae flowers were recorded. Natural arrays contained unmanipulated flowers of two species. Flowers in 'color' arrays differed only in color, while 'morphology' arrays had the effect of color eliminated. Different combinations of color and morphology preferences were found in the species examined. P. campestris and C. oetus had color but no 'morphology' preferences. S. mormonia and L. heteronea had color and morphology preferences. The behavior of L. heteronea suggests a potential to exert correlational selection on associations of flower traits. In this first field experimental test of butterfly flower constancy, no species showed constancy.