Premium
Population dynamics of wild rodents induce stochastic fadeouts of a zoonotic pathogen
Author(s) -
Guzzetta Giorgio,
Tagliapietra Valentina,
Perkins Sarah E.,
Hauffe Heidi C.,
Poletti Piero,
Merler Stefano,
Rizzoli Annapaola
Publication year - 2017
Publication title -
journal of animal ecology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.134
H-Index - 157
eISSN - 1365-2656
pISSN - 0021-8790
DOI - 10.1111/1365-2656.12653
Subject(s) - outbreak , population , biology , hantavirus , wildlife , seroprevalence , extinction probability , ecology , geography , extinction (optical mineralogy) , susceptible individual , transmission (telecommunications) , zoology , demography , population size , virology , virus , serology , immunology , paleontology , sociology , antibody , electrical engineering , engineering
Summary Stochastic processes play an important role in the infectious disease dynamics of wildlife, especially in species subject to large population oscillations. Here, we study the case of a free ranging population of yellow‐necked mice ( Apodemus flavicollis ) in northern Italy, where circulation of Dobrava‐Belgrade hantavirus (DOBV) has been detected intermittently since 2001, until an outbreak emerged in 2010. We analysed the transmission dynamics of the recent outbreak using a computational model that accounts for seasonal changes of the host population and territorial behaviour. Model parameters were informed by capture‐mark‐recapture data collected over 14 years and longitudinal seroprevalence data from 2010 to 2013. The intermittent observation of DOBV before 2010 can be interpreted as repeated stochastic fadeouts after multiple introductions of infectious rodents migrating from neighbouring areas. We estimated that only 20% of introductions in a naïve host population results in sustained transmission after 2 years, despite an effective reproduction number well above the epidemic threshold (mean 4·5, 95% credible intervals, CI: 0·65–15·8). Following the 2010 outbreak, DOBV has become endemic in the study area, but we predict a constant probability of about 4·7% per year that infection dies out, following large population drops in winter. In the absence of stochastic fadeout, viral prevalence is predicted to continue its growth to an oscillating equilibrium around a value of 24% (95% CI: 3–57). We presented an example of invasion dynamics of a zoonotic virus where stochastic fadeout have played a major role and may induce future extinction of the endemic infection.