The spatiotemporal dynamics of exposure of wild ungulates to Flavivirus shapes the patterns of West Nile virus outbreaks in Spain

Abstract

West Nile virus (WNV) is a neuropathic virus for humans, horses and birds transmitted by mosquitoes. The recent increase of clinical cases in Spain shows the need to better understand the ecology of this Flavivirus, not only in urban areas, but also where birds, WNV and mosquitoes interact with wild ungulates, to better predict/prevent WNV outbreaks. We analysed the dynamics of exposure of wild ungulates to Flavivirus along 15 years (2005-2019) in a south-to-north gradient in 5 areas where WNV cases occur since 2007. Sera of 2,418 wild ungulates (Cervus elaphus &Sus scrofa) from Doñana National Park (A1), Western Sierra Morena (A2), Central Sierra Morena (A3), Guadiana river Valley (A4) and Toledo Mounts (A5) were analysed with a blocking ELISA highly specific for WNV. Weather conditions were gathered from two climatically and environmentally contrasting areas: i) A1, a thermo-Mediterranean wetland; and ii) A4, a Mediterranean continental forest/shrubland. The risk of Flavivirus exposure of ungulates was modelled with weather covariates (average winter/spring/summer temperatures and annual/winter-spring/spring/summer rainfall) by logistic binary regression. The proportion of exposed ungulates was higher in A1 (33.5%) and A2 (35.3%) than in A3 (18.7%), A4 (20.3%) and A5 (18.7%), shaping the contrasting reported incidence of WNV outbreaks in southern (n=189) and south-central (n=2) Spain. The annual patterns of exposure of ungulates also shaped the temporal emergence of WNV outbreaks. Higher winter temperature associated positively to the risk of exposure in A1 whereas in A4 the risk increased with annual rainfall whereas it decreased with increasing spring precipitation. These preliminary results point out that whereas in wetlands, where water is abundant, smoother winter temperatures better predict WNV circulation, annual rainfall seems to be a better predictor for WNV infection risk in dry continental Mediterranean areas. Wild ungulates additionally prove to be efficient predictors for Flavivirus spatiotemporal dynamics.