Wolbachia affects mitochondrial population structure in two systems of closely related Palaearctic blue butterflies

Abstract

The bacterium Wolbachia infects many insect species and spreads by diverse vertical and horizontal means. As co-inherited organisms, these bacteria often cause problems in mitochondrial phylogeny inference. The phylogenetic relationships of many closely related Palaearctic blue butterflies (Lepidoptera: Lycaenidae: Polyommatinae) are ambiguous. We considered the patterns of Wolbachia infection and mitochondrial diversity in two systems: Aricia agestis/Aricia artaxerxes and the Pseudophilotes baton species complex. We sampled butterflies across their distribution ranges and sequenced one butterfly mitochondrial gene and two Wolbachia genes. Both butterfly systems had uninfected and infected populations, and harboured several Wolbachia strains. Wolbachia was highly prevalent in A. artaxerxes and the host's mitochondrial structure was shallow, in contrast to A. agestis. Similar bacterial alleles infected both Aricia species from nearby sites, pointing to a possible horizontal transfer. Mitochondrial history of the P. baton species complex mirrored its Wolbachia infection and not the taxonomical division. Pseudophilotes baton and P. vicrama formed a hybrid zone in Europe. Wolbachia could obscure mitochondrial history, but knowledge on the infection helps us to understand the observed patterns. Testing for Wolbachia should be routine in mitochondrial DNA studies.

Figure 1

TCS haplotype networks of (a) Aricia agestis, (b) Aricia artaxerxes and (c) Pseudophilotes baton species complex, based on COI. The colours of map points correspond to BAPS clusters (coloured network background). The shades on the maps represent the distribution of each taxon. The specimens in the networks were allocated to geographical or political regions, for readers’ orientation. The haplotypes, in which some of the individuals were inspected for Wolbachia infection, are indicated, with a different mark for each bacterial strain. These marks correspond to Fig. 3; and in (c), circle = P. baton, triangle = P. vicrama, square = P. sinaicus, star = P. b. jacuticus, hexagon = P. panoptes. Note that COI sequences not marked with any symbol were obtained in databases and hence not tested for Wolbachia. Maps were created in QGIS v. 2.18 (http://qgis.org) and the graphics was compiled in Graphic for Mac v. 3.1 (https://www.graphic.com/mac/).

Figure 2

Genetic landscapes based on COI genetic distances among individual populations of (a) Aricia agestis, (b) Aricia artaxerxes and (c) Pseudophilotes baton species complex, created by the GDisPAL function in SPADS. The distance weighting parameter a = 3 was used as a presentable visualization. The map colours represent residuals of genetic distances. Maps were created in QGIS v. 2.18 (http://qgis.org) and the graphics was compiled in Inkscape v. 1.0 (https://inkscape.org/).

Figure 3

Wolbachia infection in two groups of closely related lycaenid butterflies: (a–c) Aricia agestis/A. artaxerxes: (a) distribution map, (b) Bayesian phylogenetic tree (MrBayes) of wsp, and (c) of coxA alleles. d–f Pseudophilotes baton species complex, (d) distribution map, (e) Bayesian phylogenetic tree of wsp, and (f) of coxA alleles. For a part of the samples, only the wsp gene was obtained, such samples were categorized by their wsp alleles. Maps were created in QGIS v. 2.18 (http://qgis.org) and the graphics was compiled in Graphic for Mac v. 3.1 (https://www.graphic.com/mac/).

Figure 4

Mitochondrial Bayesian phylogenetic tree of Pseudophilotes spp. COI haplotypes. The branch labels represent posterior probabilities > 0.4. Haplotypes positive for Wolbachia are marked. The tree was visualized in FigTree v. 1.3.1 (https://github.com/rambaut/figtree). Graphics was compiled in Graphic for Mac v. 3.1 (https://www.graphic.com/mac).