Patterns and mechanisms of divergence in butterflies across spatial scales [dissertation]

Call Number:

URL:

Keywords:

Abstract:

Evolutionary divergence is caused by a great number of factors, but a key component is geographic isolation. Lineages may diverge due to selection or in the absence of selection or geographic contact, and it is often difficult to identify the relative strengths of geographic isolation vs. local adaptation, especially within cryptic species. This dissertation broadly studies evolutionary dynamics across spatial scales, ranging from the congruency of intercontinental, regional, and population differentiation to the consequences of dispersal events and secondary contact between partially diverged populations. Chapter 1 describes a comparative phylogeographic study of “weedy” Pan-American butterflies. Few comparative phylogeography studies have been conducted at very large spatial scales, encompassing species that are distributed across multiple continents. The author tested for congruence of demographic histories across a range of spatial scales and to investigate the effects of human-facilitated range expansion in several Pan-American butterfly species associated with weedy, human-modified habitats. The mtDNA cytochrome c oxidase subunit II region (COII) was sequenced for Hylephila phyleus (27 localities, n=241), Lerodea eufala (10 localities, n=93), Erynnis funeralis (4 localities, n=18), and Agraulis vanillae (13 localities, n=99) across their North and South American ranges. Data from these conspecifics was compared with homologous COII variation between allopatric congener pairs on both continents whose ranges approximate the conspecifics and also share similar weedy habitat associations: Ancyloxypha numitor (n=11) vs. A. nitedula (n=11), Vanessa annabella (n=5) vs. V. carye (n=5), and Euptoieta claudia (n=4) vs. E. hortensia (n=3). Tested for were similarities in demographic histories within and across continents for each species using pairwise distances, population genetic statistics, mismatch distributions, and deviations from mutation-drift equilibrium. Mean pairwise divergence across continents was lower for Lerodea eufala (0.05% ± 0.03 SE) and Hylephila phyleus (0.30% ± 0.11 SE) (with several shared Pan-American haplotypes each) compared to Erynnis funeralis (1.25% ± 0.37 SE) and Agraulis vanillae (3.1% ± 0.59 SE) (both with no shared haplotypes). Differentiation between congeneric species pairs was generally significantly higher than conspecific divergence across continents, but North and South American populations of Agraulis vanillae were more divergent than between Vanessa annabella and V. carye (2.4% ± 0.5 SE). Deviations from mutation-drift equilibrium were found in A. vanillae. Population-level variation was greater than the variation across continents for H. phyleus and L. eufala. Little congruence between these taxa was found across continents, although similar demographic patterns can be detected at smaller regional levels. Except for California populations of some species, the North American distributions of these weedy butterfly species appear to largely pre-date the influences of human-facilitated range expansion. Chapters 2 and 3 shift from the multi-taxon comparative focus of Chapter 1 to study the evolutionary implications of secondary contact between shallowly diverged sister lineages using the sooty hairstreak butterfly complex (Satyrium fuliginosa and S. semiluna). Secondary contact between partially divergent lineages can either reinforce or reduce pre-zygotic reproductive barriers. Satyrium semiluna and S. fuliginosa (Lycaenidae: Theclinae) are cryptic sexually monomorphic hairstreaks found in isolated subalpine and steppe habitats of western North America. Delimitation of the taxa is largely based on the relative presence and absence of the male androconial patch (the “stigma”). Androconial scales are believed to produce pheromones critical to courtship and species-level recognition and therefore should be phylogenetically conserved. In Chapter 2, evidence is found, using phylogenetic estimates based on mitochondrial DNA divergence across 36 global Satyrium, 13 Neolycaena, and four outgroup Theclinae taxa, that the absence of the stigma in S. fuliginosa is a derived loss that is independent of at least three other independent losses in Eurasia. Using sequence data from known markers and anonymous AFLP loci, the author studied the phylogeographic relationships within the entire S. fuliginosa/semiluna complex and documented a narrow hybrid zone in the California Sierra Nevada that also corresponds with populations with polymorphic wing phenotypes. All males from the population with the strongest signature of hybridization (Mt. Lincoln, Placer Co., CA) lack the stigma (therefore classified as S. fuliginosa morphologically), but the individuals are on average 2:1 genomic admixtures of S. fuliginosa and S. semiluna, and the majority (to near fixation in one population) possess the mitochondrial signatures of stigma-bearing S. semiluna. This data suggest unidirectional introgression of alleles typical of S. semiluna into S. fuliginosa. In Chapter 3, laboratory and field experiments were used to test for the presence of pre-zygotic barriers to gene flow across this Sierra Nevada hybrid zone. Individuals from all pairwise combinations males and females of several populations from either side of this contact zone courted and mate in captivity, but too few such interactions were observed for statistical confidence to explicitly test for potential assortative mating. Captive conditions cannot fully replicate the conditions under which true dispersal events between adjacent divergent populations and often under-estimate pre-zygotic barriers to gene flow. Therefore, the author conducted a reciprocal field experiment by exposing wild males at five populations from across the Sierra Nevada hybrid zone (two populations of “pure” S. fuliginosa, two populations of “pure” S. semiluna, and one hybrid population) to replicated arrays of females from each of those same five populations and recording flight deviations (a surrogate for courtship attempts). Observed was that males (regardless of taxon) generally preferred to court, or at least were neutral toward, females from S. semiluna populations under wild conditions. S. semiluna males strongly preferred S. semiluna females. Interestingly, while S. fuliginosa females did relatively better in an environment of S. fuliginosa males than in an environment of S. semiluna males, S. semiluna females were still courted at a greater level than S. fuliginosa females in an environment of S. fuliginosa males. This interaction appears to be largely driven by the fact that females from the hybrid Mt. Lincoln population (morphologically S. fuliginosa) were consistently courted by wild males at the same frequency as S. semiluna females were. The combination of genomic data and field behaviors suggests that S. fuliginosa and S. semiluna are not reproductively isolated and that S. fuliginosa may retain preferences for S. semiluna characteristics.

Notes:

ELECTRONIC FILE - Zoology: Invertebrates

SWAP (2/19/16) citation:
Runquist EB. 2012. Patterns and mechanisms of divergence in butterflies across spatial scales [dissertation]. Davis (CA): University of California, Davis. 117 p.