Gene flow and the genealogical history of Heliconius heurippa

Abstract

Background: The neotropical butterfly Heliconius heurippa has a hybrid colour pattern, which also contributes to reproductive isolation, making it a likely example of hybrid speciation. Here we used phylogenetic and coalescent-based analyses of multilocus sequence data to investigate the origin of H. heurippa.

Results: We sequenced a mitochondrial region (CoI and CoII), a sex-linked locus (Tpi) and two autosomal loci (w and sd) from H. heurippa and the putative parental species, H. cydno and H. melpomene. These were analysed in combination with data from two previously sequenced autosomal loci, Dll and Inv. H. heurippa was monophyletic at mtDNA and Tpi, but showed a shared distribution of alleles derived from both parental lineages at all four autosomal loci. Estimates of genetic differentiation showed that H. heurippa is closer to H. cydno at mtDNA and three autosomal loci, intermediate at Tpi, and closer to H. melpomene at Dll. Using coalescent simulations with the Isolation-Migration model (IM), we attempted to establish the incidence of gene flow in the origin of H. heurippa. This analysis suggested that ongoing introgression is frequent between all three species and variable in extent between loci.

Conclusion: Introgression, which is a necessary precursor of hybrid speciation, seems to have also blurred the coalescent history of these species. The origin of Heliconius heurippa may have been restricted to introgression of few colour pattern genes from H. melpomene into the H. cydno genome, with little evidence of genomic mosaicism.

Figure 1

mtDNA Phylogenetic tree. Phylogenetic relationships of H. heurippa (H) with other populations of H. melpomene and H. cydno based on CoI and CoII sequences. C and M identify H. m. melpomene and H. c. cordula individuals of the putative parental species. Sequence ID's beginning with AF and AY indicate GenBank accession numbers. Branch lengths and probability values (under branches) were estimated using Bayesian analysis and bootstrap support (over branches) derived from a Maximum Parsimony analysis. Countries of origin are identified using the following abbreviations: P = Panama and C = Colombia; Abbreviations of species names are m. = melpomene, c. = cydno.

Figure 2

Allele networks for nuclear genes. Yellow, Red and Blue are H. c. cordula, H. m. melpomene and H. heurippa alleles. Respective alleles are also identified with the letters C, M and H, followed by the individual number and allele number. Black dots are hypothetical ancestors. Sizes of the circles reflect allele frequencies in the population. (A) Tpi, (B) Dll, (C)inv, (D) w and (E) sd.