A dynamic prevalence of Francisella tularensis and Bartonella spp. was reported in irruptive common vole (Microtus arvalis) populations during 2013–2015 from agricultural landscapes of northwestern Spain (1,2). In that area, notifiable tularemia has been endemic since 1997, and human cases periodically occur during outbreaks in voles (3,4). Prevalence of F. tularensis and Bartonella spp. in voles increases with vole density (1,2), highlighting the key role of fluctuating rodents in shaping zoonoses dynamics (1–4). Rodent ectoparasites often play a major role in transmitting zoonotic pathogens. In the population studied, ticks rarely infest voles (2% prevalence), whereas fleas are much more prevalent (68%) (2). Nevertheless, any potential role for vole fleas in the circulation of F. tularensis or Bartonella spp. in natural environments remains unknown. To elucidate realistic transmission route scenarios in host-dynamic environments (5–8), we investigated whether zoonotic bacteria occur concomitantly in voles and fleas.

Our main goal was to study the prevalence of F. tularensis in fleas collected from voles previously tested for tularemia (1). We screened flea DNA in search of 6 main zoonotic bacteria simultaneously (Anaplasma phagocytophilum, Bartonella spp., Borrelia spp., Coxiella burnetii, F. tularensis, and Rickettsia spp.), following the same molecular procedure (multiplex PCR) (9) previously used to screen vole pathogens (1,2). Voles and fleas were live-trapped in northwestern Spain during March 2013–March 2015 (Appendix). We collected fleas from each individual vole and identified and grouped them in pools (pool = total fleas/vole). Three flea species parasitize common voles in the area: Ctenophthalmus apertus, Nosopsyllus fasciatus, and Leptopsylla taschenbergi (2). We screened monospecific pools (all fleas in a pool belonged to the same species and came from the same vole host), for a sample size of 90 vole hosts (pools) and 191 fleas. We screened 78 C. apertus fleas (39 pools) and 113 N. fasciatus fleas (51 pools). Among the 90 voles providing fleas, 27 were F. tularensis PCR–positive; the remaining 63 were negative (1). Of these same 90 voles, 45 were Bartonella PCR–positive and 45 were negative. Seventeen were positive for both F. tularensis and Bartonella spp. (2).

Flea pools had an average of 2.12 fleas (range 1–9); however, most (>70%) contained 1 (51%) or 2 (22%) fleas (Table). We did not detect DNA from pathogens other than F. tularensis and Bartonella spp. in fleas. Three (3%) flea pools harbored F. tularensis DNA; we estimated the overall prevalence at 6%. F. tularensis prevalence in both flea species was low (1 positive pool of 51 in N. fasciatus and 2 of 39 in C. apertus). All F. tularensis PCR–positive flea pools came from F. tularensis PCR–positive voles, and prevalence of F. tularensis in fleas was significantly associated with its prevalence in voles (analysis of variance [ANOVA], R² = 0.072, F_{0.05, 1, 88} = 6.81; p = 0.011). Of note, all fleas containing F. tularensis DNA were collected during July 2014, when vole populations reached top densities and tularemia prevalence peaked among them (33%) (1). The low prevalence of F. tularensis detected in fleas carried by infected hosts (3 of 27 pools) and the detection of infected flea pools only when abundance of the bacterium in the environment was highest (during vole peaks) (1,4) suggest that the quantitative role of fleas in the circulation of F. tularensis might be modest.

Conversely, the role of fleas in the circulation of Bartonella spp. seems much more relevant. We detected Bartonella spp. in 28 (37%) flea pools and in both flea species (37% of N. fasciatus and 23% of C. apertus) (Table). We detected Bartonella spp. in fleas collected from Bartonella PCR–positive and Bartonella PCR–negative voles in nearly equal proportions (51% vs. 44%) (Table). The average prevalence of Bartonella spp. in fleas was not associated with its prevalence in voles (ANOVA, R² = 0.006, F_{0.05, 1, 88} = 0.53; p = 0.467). We found a higher Bartonella spp. prevalence in N. fasciatus (65%) than in C. apertus (33%). We identified 3 Bartonella species among fleas (B. taylorii [17%], B. grahamii [14%], and B. rochalimae [3%]), as well as mixed infections (Appendix). These findings are in accordance with other research showing fleas as a main vector of Bartonella spp. (5). Although F. tularensis and Bartonella spp. have been simultaneously detected in ≈13% of voles during population density peaks (2), we identified no co-infection among flea pools (ANOVA, R² = 0.011, F_{0.05, 1, 88} = 0.97; p = 0.328).

Our data show that F. tularensis and Bartonella spp. occur in the fleas infesting wild common voles in northwestern Spain, with notable differences in prevalence (6% and 51%, respectively) and associations with prevalence in vole hosts. Future studies are needed to determine the role of fleas in the circulation of these pathogens in nature and in particular to ascertain any effective vectoring of F. tularensis.

Top