**4. Discussion**

Evaluating relationships between zoonotic pathogens and reservoir hosts, particularly in the context of host immune defenses, can provide new insights into these natural systems and the risks posed to humans. The immunological phenotype of hosts is shaped by both parasite pressure and parasite load [1]. To help to disentangle these two factors, we quantified immunological and related physiological indices and parasite load parameters in bank voles and wood mice from wooded sites in the Netherlands varying in parasite pressure. We found immunological and physiological differences between groups that were and were not infected or infested with parasites, but we also found differences among sites. Defining the biological meaning of such differences, even when using some of the most widely applied assays of ecological immunology, remains challenging, but our results offer some new context to the measured indices. For example, interspecific differences in hemolysis were larger than intraspecific differences related to infestation (parasite load), supporting an earlier idea that this index is relatively invariable in relation to current health status, despite the taxonomic variation it displays [7,9]. In addition, both infestation (parasite load) and site (parasite pressure) associated differences were related to haptoglobin values outside of clinically normal ranges (rodent: 0.25–0.51 mg mL−1, murine: 0.00–0.10 mg mL−1; both reported by the assay manufacturer). Thus, individuals of both rodent species that were either feeding *I. trianguliceps* ticks or living at Herperduin can be viewed as enduring a type of systemic innate immune response, known as an acute phase response [42]). Overall, we can safely conclude that the measured immunological defenses are shaped by a combination of current parasite load and differences in parasite pressure among spatially distinct sites.

We hypothesized that if parasite pressure drives immunological phenotype, then populations from sites that differ in parasite pressure should express different levels of the measured immune indices. In general, higher parasite pressure is thought to select for stronger immune systems [1]. We found partial support for our hypothesis in the form of di fferences among sites in terms of haptoglobin, hemagglutination, and body-mass corrected spleen size. In all three cases, one site di ffered from the other three. This suggests that circumstances specific to some sites might influence parasite pressure (e.g., which parasite species and pathogen strains successfully colonized a site, the overall parasite and pathogen richness or diversity), evolutionary responses to that pressure (e.g., host population genetics), or both. However, the patterns reported here are inconsistent with previously reported di fferences in parasite pressure, which was characterized by densities of *Ixodes ricinus* and prevalences of several tick-borne pathogens [15]. For example, sites with higher levels of haptoglobin (Herperduin), hemagglutination (Buunderkamp, Stameren and Herperduin), and body-mass corrected spleen size (Stameren) were not the same as the site with the highest parasite pressure in 2013–2014 (Buunderkamp). We did find that endoparasite loads were highest in rodents from Herperduin, which could explain the higher level of haptoglobin at this site, suggesting that a more thorough approach is needed to quantify parasite pressure at a site [2]. Even though sites were on average 30 km apart and were previously shown to have clear di fferences in tick densities and tick-borne microparasite prevalences [15], these distinctions might not have consistently translated to di fferent immunological phenotypes in measurable ways. Previous studies showing between-site di fferences in host immunology employed study designs with extreme environmental di fferences [43].

We also found support for our second hypothesis: individuals burdened with higher parasite loads were expected to di ffer from those carrying lower loads in terms of the measured immune indices. The direction of e ffect likely depends on the parasite type (immunostimulatory vs. immunosuppressive) and immune index under consideration. Indeed, we found that the highest and lowest levels of haptoglobin and the neutrophil to lymphocyte ratio were associated with infection or infestation. Concentrations of haptoglobin were higher in individuals infested with the tick *Ixodes trianguliceps*. Haptoglobin is a positive acute phase protein [44], meaning concentrations increase in response to infection, inflammation, or trauma [42], any of which could result from the mouthparts of a tick puncturing the skin of a host and the multiday feeding period that follows [45]. We also found the highest neutrophil to lymphocyte ratios in animals infected with *B. miyamotoi*. This result could be caused by increased neutrophil production or decreased lymphocyte numbers in these infected animals, but neither of these e ffects were seen in a case-study of a human patient [46]. Furthermore, we found the lowest neutrophil to lymphocyte ratios in animals infested with *Laelaps agilis*. This species of mite is known to be a vector of *Hepatozoon* spp., blood parasites that can infect wild rodents, but for which we did not test our samples [47]. Since *Hepatozoon* spp. can infect leukocytes in rodents [48,49], such an infection might be a mechanism behind the mite infestation e ffect; however, in other animals (e.g., frogs [50]), neutrophil to lymphocyte ratio was not correlated with infection intensity of *Hepatozoon* spp. Nevertheless, since increased neutrophil to lymphocyte ratios are often associated with stressed or diseased states [51], the lower values we observed in mite infested individuals sugges<sup>t</sup> another process at work (e.g., possible immunosuppression).

Finally, we found di fferences related to species identity and body mass. Hemolysis was the only immune index for which we found a di fference between the two study species: Overall bank voles exhibited higher titers than wood mice. Furthermore and tangentially related to our a priori expectation, hemolysis was the only immune index for which we found an infection-related di fference between the two study species. Hemolysis titers were higher in *Heligmosomoides polygyrus* (helminth) infected wood mice but not in similarly infected bank voles. *Heligmosomoides polygyrus*, a commonly used model of helminth infections, is known to regulate immune function in laboratory mice [52]. To our knowledge, our study is the first to show immunological di fferences associated to natural infection with *H. polygyrus*. It must be noted, however, that the e ffects of microparasite infection status never showed this type of species dependence. We also found that four immune indices showed three di fferent types of relationships with body mass. Hemagglutination and body-mass corrected spleen size related positively with body mass; haptoglobin related negatively with body mass; and hematocrit showed an optimum at middle body masses. While the e ffects of age on immune

function in rodents are known, if not fully understood [2], our results highlight the complexity of these dynamics. The immune system in wild rodents does not simply "mature" as individuals grow heavier (and older), even if some individual components, such as hemagglutination, show such a pattern here and elsewhere [7,9]. Instead, the observed relationships hint at an influence of body condition or composition, but this potential mechanism could not be investigated given our lack of data on the structural size of individuals [18].

Overall, our findings o ffer new insights into relationships between specific parasites and immunological and physiological indices, as well as broader di fferences between species and among sites. Notably, di fferences related to study site and infection status seemed to exert a greater impact on immune phenotype than host species identity, even though the two rodent study species are thought to play di fferent roles in the maintenance and transmission of tick-borne zoonotic pathogens [6,16]. While our results document influential roles for both the environmental characteristic of parasite pressure and the organismal characteristic of parasite load on the immunological phenotype of wild animals, additional studies linking disease ecology and ecological immunology are needed to better understand the complexities of how host–parasite interactions play out through space and time in di fferent environments. For example, repeated sampling of sites over longer (i.e., multiyear) time periods would be invaluable for characterizing the potential for parasite pressure to drive immune system evolution. Lastly, our study illustrates the possibility and added value of using a holistic approach targeting diverse parasites and multiple aspects of host immunology and physiology when investigating zoonotic pathogens and their vectors and reservoirs [53]. The resulting insights, such as overall di fferences between species (e.g., hemolysis) and interspecific similarities and di fferences in the immunomodulatory e ffects of infection (e.g., neutrophil to lymphocyte ratio and hemolysis, respectively), can help to shape new questions, for example, about host competence [53]. In this study, our focus was on understanding immunological variation, but a similar analytical approach could be used to explore variation in the burden (e.g., infection intensity) of one or more parasites.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2076-2615/9/12/1015/s1, Supplementary Table S1: Site characteristics, and Supplementary Table S2: Full dataset. Supplementary Figure S1: Map with locations of study sites. Left: map of the Netherlands with location of detailed map. Right: detailed map with location of the four study sites.

**Author Contributions:** Conceptualization, T.R.H., E.J.B., B.H., H.S., and K.D.M.; methodology, T.R.H., E.J.B., B.H., M.M., F.F.J.F., H.S., and K.D.M.; data analysis, T.R.H., E.J.B., and B.H.; data curation, T.R.H.; writing—original draft preparation, E.J.B. and B.H.; writing—final draft preparation, T.R.H. and K.D.M.; writing—review and editing, T.R.H., E.J.B., B.H., M.M., F.F.J.F., H.S., and K.D.M.; data visualization, T.R.H.; supervision, T.R.H. and K.D.M.; project administration, T.R.H. and K.D.M.; funding acquisition, T.R.H., M.M., F.F.J.F, H.S., and K.D.M.

**Funding:** This research was funded by the de Vos Vector Borne Diseases Fund at Wageningen University and by the Dutch Ministry of Health, Welfare, and Sport (VWS).

**Acknowledgments:** We are grateful to Herman J.W.M. Cremers (Utrecht University) for his help with ectoparasite species identification and to Marieke de Swart (National Institute for Public Health and Environment) for her help with microparasite screening. We also thank Ger de Vries Reilingh, Joop A.J. Arts, Rudie E. Koopmanschap, Henk K. Parmentier and the rest of the Adaptation Physiology Group (Wageningen University) for granting us access to their laboratory and for technical assistance.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.
