Endemic pathogens, such as hantaviruses, have been hypothesized to cause no apparent symptoms or fitness costs to their hosts. However, we found PUUV infection to decrease the over-winter survival of wild bank voles. We also found wild bank voles to shed PUUV via urine, faeces, and saliva throughout their life span without any remarkable decline, in contrast to earlier results from laboratory-reared rodent hosts.

For the first time, dynamics of PUUV infection were studied during winter, when the majority of NE cases occur in the boreal zone. We found PUUV-infected bank voles to be most abundant in the winters of increase and peak years of the 3-year density cycle. In bank voles, the prevalence of PUUV infection showed a regular, seasonal fluctuation, which resulted from seasonal population turnover and the positive correlation between age and the likelihood of being PUUV infected. However, despite its regular fluctuation, PUUV prevalence in voles is not a good predictor of human NE risk since the periods of high prevalence coincided with low NE incidence in humans.

Aggression has been suggested as the key driver for other hantaviruses in their host species, but the rate of PUUV transmission in bank voles was higher outside the breeding season, when bank voles do not show aggressive behaviour, than during the breeding season. The high rate of transmission outside the breeding season may be attributed to subnivean conditions that promote virus stability, lower immune response during cold conditions, and high host density in fall.

We also found evidence for the “dilution effect” hypothesis, which assumes non-host species to reduce virus transmission between hosts: the prevalence of PUUV was low in bank voles when other small mammals were abundant.

Male sex bias in infection is a general phenomenon that has also been observed in several hantavirus-host systems. We found a male bias in PUUV infection only in overwintered, breeding bank voles, whereas a female bias was seen in summer-born breeding animals. In non-breeding animals, no sex differences existed. Therefore, the effects of host sex in hantavirus transmission may be more complex than previously thought.

Forest habitats disturbed by intensive forest management were associated with a higher likelihood of PUUV infection in bank voles. This finding could be explained by the poorer quality of these habitats, leading to lower condition and higher susceptibility, and also by more favourable environmental conditions for virus survival outside the host. Despite the higher infection prevalence in voles, the total number of PUUV-infected bank voles was 46-64% lower in young, intensively managed than in undisturbed, old forests. Thus, environmental change per se does not automatically lead to relative success of species that serve as reservoirs for zoonotic pathogens, and thereby, to increased human disease risk.

The results of this thesis encourage further studies of host-pathogen interactions in natural conditions, and in different host-hantavirus systems. They also provide a framework for risk models aiming at reduction of human hantavirus infections.