Range-wide genetic analysis of Dermacentor variabilis and its Francisella-like endosymbionts demonstrates phylogeographic concordance between both taxa

Emily L. Kaufman, Nathan E. Stone, Glen A. Scoles, Crystal M. Hepp, Joseph D. Busch, David M. Wagner

The American dog tick, Dermacentor variabilis, is an important vector of pathogens to humans, wildlife and domestic animals in North America. Although this tick species is widely distributed in the USA and Canada, knowledge of its range-wide phylogeographic patterns remains incomplete. We carried out a phylogenetic analysis of D. variabilis using samples collected from 26 USA states and five Canadian provinces. Tick samples (n = 1053 in total) originated from two main sources: existing archives (2000-2011), and new collections made from 2012 to 2013. We sequenced a 691 bp fragment of the cox1 gene from a subset (n = 332) of geographically diverse D. variabilis. DNA extracted from individual ticks (n = 1053) was also screened for a Francisella-like endosymbiont, using a targeted 16S rRNA sequencing approach, and important pathogens (Rickettsia spp. and Coxiella burnetii), using species-specific quantitative PCR assays. Maximum parsimony analysis of cox1 sequences revealed two major groups within D. variabilis with distinct geographical distributions: one from the eastern USA/Canada (Group 1) and one from the west coast states of the USA (California and Washington; Group 2). However, genetic subdivisions within both of these two major groups were weak to moderate and not tightly correlated with geography. We found molecular signatures consistent with Francisella-like endosymbionts in 257 of the DNA extracts from the 1053 individual ticks, as well as Rickettsia spp. and Coxiella burnetii in a small number of ticks (n = 29 and 2, respectively). Phylogenetic patterns for Francisella-like endosymbionts, constructed using sequence data from the bacterial 16S rRNA locus, were similar to those for D. variabilis, with two major groups that had a nearly perfect one-to-one correlation with the two major groups within D. variabilis. Our findings reveal a distinct phylogenetic split between the two major D. variabilis populations. However, high levels of genetic mixture among widely separated geographical localities occur within each of these two major groups. Furthermore, our phylogenetic analyses provide evidence of long-term tick-symbiont co-evolution. This work has implications for understanding the dispersal and evolutionary ecology of D. variabilis and associated vector-borne diseases.