Date:
Location:
Title: Analyzing the evolution and diversification of Lepidoptera using multi-spectral images
Abstract: Insects comprise an estimated 60% of all described species, and butterflies are among the best known of all insect taxa. The order Lepidoptera (butterflies and moths) contains approximately 160,000 species in 126 families. Although their diverse colors and shapes have been investigated independently in many specific ecological and evolutionary studies, few studies have surveyed changes in shape and color across the entire evolutionary history of the group. To be able to tackle this question in a more systematic way, I developed an economical imaging system with integrated hardware and software to capture multispectral images of Lepidoptera with high efficiency (Chapter 1). This method facilitates the comparison of colors and shapes on both the dorsal and ventral sides of species at fine and broad taxonomic scales. Unlike conventional imaging, it takes into account a broader spectrum of wavelengths in the electromagnetic spectrum, with bands ranging from the UV through to the near infrared, thereby including all wavelengths that Lepidoptera (many of which are tetrachromatic) are known to perceive.
In Chapter 2, I used comparative methods to analyze the evolution of wing shape and size for 2,388 specimens representing 190 exemplar species included in a recent tribal-level phylogeny of the butterflies (Papilionoidea), a monophyletic group within the Lepidoptera that includes 7 families, ~1,815 genera, and ~18,728 described species. Different wing shapes result in differences in flight performance, which is related to the probability of escaping from predators and unfavorable environmental conditions. Results show that rather than shape, butterfly size and overlap of fore- and hind- wings largely determine flight performance, and that different flight systems can be identified based on parameters of wing size and shape.
In Chapter 3, I developed a heuristic model of lepidopteran vision that generates several hypotheses regarding eye structure and function that can be assessed using empirical data. A survey of 2,730 specimens representing 224 species across butterflies and moths indicates that the advantages of binocular vision select for morphological architecture enhancing detectability and stereopsis in nocturnal and diurnal species, respectively.
In Chapter 4, I evaluated the ecological consequences of urbanization, using a structural equation model that incorporates urban land use, remotely sensed environmental data, phenotypic trait variation, and biodiversity of Geometridae in New England, with a focus on regional collections of the peppered moth, Biston betularia. Our results support multiple hypotheses: changes in land type, temperature, and vegetation were all significantly correlated with specimen brightness, while air pollution and nightlight were not. These results were contingent on spatial, temporal, and spatiotemporal contexts.
Committee: Naomi Pierce (Advisor), Andrew Davies, David Haig (Chair), Paul Moorcroft