Diversification, selective sweep, and body size in the invasive Palearctic alfalfa weevil infected with Wolbachia

Abstract

The alfalfa weevil Hypera postica, native to the Western Palearctic, is an invasive legume pest with two divergent mitochondrial clades in its invading regions, the Western clade and the Eastern/Egyptian clade. However, knowledge regarding the native populations is limited. The Western clade is infected with the endosymbiotic bacteria Wolbachia that cause cytoplasmic incompatibility in host weevils. Our aim was to elucidate the spatial genetic structure of this insect and the effect of Wolbachia on its population diversity. We analyzed two mitochondrial and two nuclear genes of the weevil from its native ranges. The Western clade was distributed in western/central Europe, whereas the Eastern/Egyptian clade was distributed from the Mediterranean basin to central Asia. Intermediate mitotypes were found from the Balkans to central Asia. Most Western clade individuals in western Europe were infected with an identical Wolbachia strain. Mitochondrial genetic diversity of the infected individuals was minimal. The infected clades demonstrated a higher nonsynonymous/synonymous substitution rate ratio than the uninfected clades, suggesting a higher fixation of nonsynonymous mutations due to a selective sweep by Wolbachia. Trans-Mediterranean and within-European dispersal routes were supported. We suggest that the ancestral populations diversified by geographic isolation due to glaciations and that the diversity was reduced in the west by a recent Wolbachia-driven sweep(s). The intermediate clade exhibited a body size and host plant that differed from the other clades. Pros and cons of the possible use of infected-clade males to control uninfected populations are discussed.

Figure 1

Statistical parsimony network of mitochondrial COI-tRNA^Leu-COII and Cyt b-tRNA^Ser-ND1 of Hypera postica in its native range. Generated using TCS 1.21.

Figure 2

Geographic distribution of mitochondrial clades of Hypera postica in its native range. Pie chart sizes for clades are proportional to sample sizes. Wolbachia infection (dark gray ring: infected; white (partial) ring: uninfected) is only indicated for the Western clade individuals because none of the Egyptian/Eastern clade or the intermediate clade were infected. The background map was obtained from Fotolla, https://stock.adobe.com/jp/photos/.

Figure 3

Statistical parsimony network of nuclear 28S and EF-1α. See Fig. 1 for colors for regions. Generated using TCS 1.21.

Figure 4

Bayesian consensus tree of Wolbachia strain based on ftsZ, coxA, and hcpA. Strain codes, if available, followed by host species and Wolbachia supergroups are shown. The Wolbachia strain, wHypera1, infecting Hypera postica in its native range, is shown in bold. Bayesian support values (posterior probabilities > 0.7) are shown near nodes. The outgroup is Anaplasma marginale (Alphaproteobacteria: Rickettsiales: Anaplasmataceae). Generated using MrBayes 3.2.6. Host strains, Wolbachia isolates and GenBank accession numbers are listed in Supplementary Table S3.

Figure 5

(a) The maximum clade credibility (MCC) tree of Hypera postica resulting from a discrete Bayesian phylogeographic analysis, based on one sample per clade per country. The outgroup is H. miles. The nodes and branches are colored according to the most probable region (left panel; see Fig. 1 for colors indicating geographic regions and Table 1 for the country codes) and Wolbachia infection state (right panel; red line: infected). The numbers on the nodes indicate posterior probabilities of the most likely ancestral states [geographic region (left panel) or Wolbachia infection state (right panel)]. Generated using BEAST 1.10.4 and visualized using FigTree 1.4.4 (http://tree.bio.ed.ac.uk/software/figtree/). (b) Map of colonization routes of H. postica. Thick lines: supported by BF (Bayes factor) > 7.0. Generated using SpreaD3 v0.9.7.1rc and visualized using Google Earth 7.3.0 (https://www.google.com/earth/).