Prevalence of Puumala virus (PUUV) in bank voles (Myodes glareolus) after a major boreal forest fire

Abstract: Our knowledge on the impact of forest fires on the prevalence (proportion of infected individuals in a population) and dynamics of zoonotic pathogens is largely limited. A large forest fire in late 2006 at Bodträskfors in northern Sweden provided a unique opportunity to investigate the effect of habitat change on disease prevalence. Pummala virus (PUUV) is one of the most prominent zoonotic viruses in this northern boreal forests with bank vole as its only competent host. Human’s infection occurs by breathing the aerosolized viral particles shed through saliva, urine and feces of the infected host. The infection causes Nephropathia Epidemia, a milder form of hemorrhagic fever with renal syndrome. The disease has relatively low death rates but can cause lifelong symptoms in humans. Here I have investigated the prevalence of PUUV in bank voles between spring and autumn of 2007-2010 and again in 2015. Small mammals were trapped in trapping plots in the Bodträskfors forest fire area (n=7), mature reference forests (n=7) and unburned clear-cuts (3). In total 1048 small mammals were trapped from which 1013 bank voles were autopsied and analyzed for anti PUUV antibody with indirect Enzyme Linked Immunosorbent Assay. I used generalized linear mixed effect model to compare PUUV prevalence in the three areas. All bank voles were also weighed and probable weight’s correlation with PUUV prevalence was investigated using nominal logistic fit and univariate ANOVA (analysis of variance). Species composition was one of the most striking results of this study. In the burned area, there appeared to be a one species system, comprised of bank voles only, between 2008-2010 and again in 2015. My results suggest a staggering 78 and 73 percent infection prevalence in burned forest in 2007 and 2015 compared to respective 55 and 44 percent infection prevalence in mature forest. This significant difference was reversed in 2010 with the reference area having the highest infection prevalence (65 to 33 percent respectively). The low species diversity, along with habitat loss due to direct effect of forest fire are suggested to be the two major contributing factors that have led to the very high infection prevalence in forest fire area. The weight of bank voles was directly correlated with infection prevalence. The weight was highest in spring, in all locations. The burned area consistently had the highest weight average in spring with the reference sites and clear-cut following it respectively. In autumn however, the weight varied slightly between areas without any consistency. PUUV prevalence differed between the burned and mature reference forests. To pinpoint the exact environmental factors that have resulted in this variation requires further environmental studies, which were out of the scope of this study. The one species system in the forest fire area along with the described infection prevalence portray a unique opportunity for identifying the environment’s effect on infection prevalence and also the epidemiologic base of infection prevalence in bank voles with regards to species diversity.