Deep mitochondrial divergence within a Heliconius butterfly species is not explained by cryptic speciation or endosymbiotic bacteria

Abstract

<h4>Background</h4>Cryptic population structure can be an indicator of incipient speciation or historical processes. We investigated a previously documented deep break in the mitochondrial haplotypes of Heliconius erato chestertonii to explore the possibility of cryptic speciation, and also the possible presence of endosymbiont bacteria that might drive mitochondrial population structure.<h4>Results</h4>Among a sample of 315 individuals from 16 populations of western Colombia, two principal mtDNA clades were detected with 2.15% divergence and we confirmed this structure was weakly associated with geography. The first mtDNA clade included 87% of individuals from northern populations and was the sister group of H. erato members of Andes western, while the second clade contained most individuals from southern populations (78%), which shared haplotypes with an Ecuadorian race of H. erato. In contrast, analysis using AFLP markers showed H. e. chestertonii to be a genetically homogeneous species with no association between mitochondrial divergence and AFLP structure. The lack of congruence between molecular markers suggests that cryptic speciation is not a plausible explanation for the deep mitochondrial divergence in H. e chestertonii. We also carried out the first tests for the presence of endosymbiontic bacteria in Heliconius, and identified two distinct lineages of Wolbachia within H. e. chestertonii. However, neither of the principal mitochondrial clades of H. e. chestertonii was directly associated with the patterns of infection.<h4>Conclusions</h4>We conclude that historical demographic processes are the most likely explanation for the high mitochondrial differentiation in H. e. chestertonii, perhaps due to gene flow between Cauca valley H. e. chestertonii and west Pacific slope populations of H. erato.