Host aggregation and local density corresponds to heterogeneity in tick-borne and rodent-borne diseases

Host-parasite-pathogen dynamics in nature have been linked to a range of extrinsic and intrinsic factors, which shape the spatio-temporal heterogeneity of environment. Here we evaluated how anthropogenic food manipulation may affect hosts local density and space-use patterns, parasite burden, and pathogen prevalence The study was carried out from February 2019 to April 2021 in Cembra Valley (Trentino, Italian Alps), in a treatment-control field experiments. Treatment sites were at ungulate feeding sites dispensing supplemental ad libitum food, where roe deer were being captured and marked with GPS collars, whereas control sites were placed at least 500 m away from the closest feeding site. In both treatment and control sites, we performed monthly live capture-mark recapture of rodents by individually marking captured individuals with Passive-Integrated Transponder (PIT) tag. Standard capture information, life-history traits and parasites’ presence were recorded, while biological samples (blood, tissue, faeces) were collected. We applied Spatially Explicit-Capture-Recapture models to spatially predict rodent density and space-use patterns, while we relied on GPS telemetry data to perform a recursion analysis with the aim of assessing roe deer revisitation patterns at treatment and control sites. Through Generalized Linear Mixed Models, we modelled how mice tick burden varied in dependence on the presence of ungulate feeding sites, rodent density and phenotypic traits (sex, status and body mass). Additionally, serological assays and molecular PCR-based methods coupled with sequencing were performed to assess the prevalence of rodent- and vector-borne pathogens at control and treatment sites. 398 yellow-necked and wood mice (Apodemus spp.), 109 bank voles (Myodes glareolus) and 12 roe deer individuals were captured at treatment sites. The availability of supplemental food shrunk rodents’ space-use, increasing mice but not voles local densities. Roe deer revisitation rate was significantly higher at treatment respect to control sites. 1790 feeding Ixodidae ticks were counted on rodents. Tick burden was significantly higher in heavier individuals, but did not depend on site. Conversely, spatial pattern of rodent- (Dobrava-Belgrade and Puumula viruses) and vector-borne (Tick-borne Encephalitis virus, Borrelia afzelii, Babesia microti, Anaplasma phagocytophylum and Hepatozoon spp.) pathogens occurrence varied among control and treatment sites. Vector-borne pathogens showed a lower prevalence at feeding sites (mean prevalence, 8.69%) than at control sites (13.89%), while the opposite was found for rodent-borne ones (1.55% and 1.16%, respectively). Our findings suggest that both woodland rodents and roe deer were attracted by supplemental food resources. We argue that the co-occurrence of incompetent (roe deer) and competent hosts (rodents) at these sites might affect infectious pathogens’ spreading. These results underline the importance of considering spatial distribution of competent and incompetent hosts, and not only overall density, to investigate zoonotic risk