doublesex is a mimicry supergene.

One of the most striking examples of sexual dimorphism is sex-limited mimicry in butterflies, a phenomenon in which one sex--usually the female--mimics a toxic model species, whereas the other sex displays a different wing pattern. Sex-limited mimicry is phylogenetically widespread in the swallowtail butterfly genus Papilio, in which it is often associated with female mimetic polymorphism. In multiple polymorphic species, the entire wing pattern phenotype is controlled by a single Mendelian 'supergene'. Although theoretical work has explored the evolutionary dynamics of supergene mimicry, there are almost no empirical data that address the critical issue of what a mimicry supergene actually is at a functional level. Using an integrative approach combining genetic and association mapping, transcriptome and genome sequencing, and gene expression analyses, we show that a single gene, doublesex, controls supergene mimicry in Papilio polytes. This is in contrast to the long-held view that supergenes are likely to be controlled by a tightly linked cluster of loci. Analysis of gene expression and DNA sequence variation indicates that isoform expression differences contribute to the functional differences between dsx mimicry alleles, and protein sequence evolution may also have a role. Our results combine elements from different hypotheses for the identity of supergenes, showing that a single gene can switch the entire wing pattern among mimicry phenotypes but may require multiple, tightly linked mutations to do so.

Cryptic genetic and wing pattern diversity in a mimetic Heliconius butterfly.

Despite rampant colour pattern diversity in South America, Heliconius erato exhibits a 'postman' wing pattern throughout most of Central America. We examined genetic variation across the range of H. erato, including dense sampling in Central America, and discovered a deep genetic break, centred on the mountain range that runs through Costa Rica. This break is characterized by a novel mitochondrial lineage, which is nearly fixed in northern Central America, that branches basal to all previously described mitochondrial diversity in the species. Strong genetic differentiation also appears in Z-linked and autosomal markers, and it is further associated with a distinct, but subtle, shift in wing pattern phenotype. Comparison of clines in wing phenotype, mtDNA and nuclear markers indicate they are all centred on the mountains dividing Costa Rica, but that cline width differs among data sets. Phylogeographical analyses, accounting for this new diversity, rewrite our understanding of mimicry evolution in this system. For instance, these results suggest that H. erato originated west of the Andes, perhaps in Central America, and as many as 1 million years before its co-mimic, H. melpomene. Overall our data indicate that neutral genetic markers and colour pattern loci are congruent and converge on the same hypothesis-H. erato originated in northwest South America or Central America with a 'postman' phenotype and then radiated into the wealth of colour patterns present today.

Comparative population genetics of a mimicry locus among hybridizing Heliconius butterfly species.

The comimetic Heliconius butterfly species pair, H. erato and H. melpomene, appear to use a conserved Mendelian switch locus to generate their matching red wing patterns. Here we investigate whether H. cydno and H. pachinus, species closely related to H. melpomene, use this same switch locus to generate their highly divergent red and brown color pattern elements. Using an F2 intercross between H. cydno and H. pachinus, we first map the genomic positions of two novel red/brown wing pattern elements; the G locus, which controls the presence of red vs brown at the base of the ventral wings, and the Br locus, which controls the presence vs absence of a brown oval pattern on the ventral hind wing. The results reveal that the G locus is tightly linked to markers in the genomic interval that controls red wing pattern elements of H. erato and H. melpomene. Br is on the same linkage group but approximately 26 cM away. Next, we analyze fine-scale patterns of genetic differentiation and linkage disequilibrium throughout the G locus candidate interval in H. cydno, H. pachinus and H. melpomene, and find evidence for elevated differentiation between H. cydno and H. pachinus, but no localized signature of association. Overall, these results indicate that the G locus maps to the same interval as the locus controlling red patterning in H. melpomene and H. erato. This, in turn, suggests that the genes controlling red pattern elements may be homologous across Heliconius, supporting the hypothesis that Heliconius butterflies use a limited suite of conserved genetic switch loci to generate both convergent and divergent wing patterns.

Does male preference play a role in maintaining female limited polymorphism in a Batesian mimetic butterfly?

Female-limited polymorphism occurs in multiple butterfly species with Batesian mimicry. While frequency-dependent selection is often argued as the driving force behind polymorphism in Batesian mimicry systems, male preference and alternative female mating strategies may also influence the maintenance of multiple female forms. Through a series of behavioural assays with the female-limited Batesian mimetic butterfly Papilio polytes, we show that males prefer stationary mimetic females over stationary non-mimetic females, but weigh female activity levels more heavily than female wing pattern when choosing between active mimetic and active non-mimetic females. Male preference for mimetic vs. non-mimetic females is independent of male genotype at the locus responsible for the female wing pattern, the autosomal gene doublesex. However male genotype does influence their response to active females. Male emphasis on female behaviour instead of appearance may reduce sexual selection pressures on female morphology, thereby facilitating frequency-dependent natural selection due to predation risk and toxic model abundance.

Reinforcement of mate preference among hybridizing Heliconius butterflies.

Recent models of mate preference evolution suggest that direct selection on alleles at preference loci and correlated evolution of preference with locally adapted mating cues are more likely to drive the evolution of assortative mate preference than reinforcement. Mate preference evolution in mimetic Heliconius butterflies has been attributed to all three forms of selection, but here we show that reinforcement has been critical. By examining geographical variation in assortative mating and male mate preference among seven populations of three hybridizing Heliconius species from Costa Rica, we found pronounced character displacement of preference such that sexual isolation was enhanced in areas of interspecific contact. Of the different explanations for the evolution of assortative mate preference, only reinforcement is dependent on interspecific contact in this system. Thus, the observed pattern of reproductive character displacement of mate preference is best explained as a product of indirect selection generated by natural selection against nonmimetic hybrids.

Lack of genetic differentiation among widely spaced subpopulations of a butterfly with home range behaviour.

We examine seven geographically separate subpopulations of Heliconius charithonia, a butterfly with well-documented home range behaviour, in Miami-Dade County, Florida, for genetic differentiation using cellulose acetate electrophoresis. These subpopulations exhibit little genetic variation (percent polymorphic loci = 27, average heterozygosity = 0.103) especially in comparison to populations of the same and related species from mainland South America. Allele frequencies do not differ among the subpopulations in south Florida and estimates of Wright's fixation index (FST) support that there is no detectable genetic differentiation among them. This result supports an earlier finding that the dispersal ability of Heliconius butterflies may be underestimated. However, it is unlikely that increased dispersal ability alone could account for the lack of genetic differentiation observed among subpopulations separated by almost 80 km. Given the likely effective population size of these subpopulations (Ne = 205) and the average generation time of this species in the subtropics (in the range of 30-90 days), this lack of genetic differentiation is best explained by current or very recent gene flow following a stepping-stone model. Furthermore, this result provides evidence that the current extensive degree of habitat fragmentation surrounding the city of Miami does not limit gene flow among urban subpopulations of Heliconius charithonia.