Phylogeny and biogeography of Heliconius butterflies inferred from mitochondrial DNA sequences

The Heliconius butterflies and close relatives (Lepidoptera: Nymphalidae: Heliconiinae) have been an important model system for understanding the evolution and ecological genetics of Mullerian mimicry. They have also played a crucial role in the debate over hostplant-herbivore coevolution. This work encompasses two levels of analysis. Chapters one and two assess phylogenetic relationships of Heliconius and Eueides species, employing a new, independent character system (mtDNA sequences). A number of major differences occur between the traditional phylogeny of Heliconius and this new hypothesis of relationships. Most dramatically, the mtDNA data indicate that Heliconius is paraphyletic with respect to Eueides, which is the sister taxon of H. charithonia. The results of the molecular analyses are compared to prior hypotheses of relationships based on phenotypic traits, and the evolution of particular morphological and behavioral characters is discussed in light of the new data. Chapters three and four focus on intraspecific relationships of two species, Heliconius erato and H. melpomene. These two species are Mullerian mimics, and exhibit tremendous parallel intraspecific geographical polymorphism. My results suggest that biogeography, and not selection for preservation of ancestral wing patterns, has played the main role in the distribution of different racial phenotypes. Similar races east and west of the Andes are not as closely related as very different looking, adjacent races within one or the other of these biogeographical regions. Wing patterns may be extremely labile genetically, and prone to convergence within as well as between species. I present a new estimate of a molecular clock for arthropod mitochondrial DNA, which suggests that the east-west split in both species occurred 1.5 to 2 million years ago, and racial differentiation within each legion began 150,000 to 200,000 years ago. These dates are compatible with the idea that races formed during cooler and drier climatic periods of the Pleistocene, when they were restricted to reduced areas of moist forest (Pleistocene refugia). However, genetic differentiation is too low in the genes examined to determine if refugia actually existed. Differentiation across extant biogeographical barriers (rivers and mountains) appears to explain most, but not all of the phenotypic variation observed.